THE COPPERBELT UNIVERSITY

School of mines and minerals sciences

CE 560 / MT 580
Process control

Lecture 16
Solution of process models using Laplace Transforms

C. Botha (Mr.)
Contacts: Chemical Engineering Department
Email: [email protected]
Alt. Email: [email protected]

May, 2023
 Solution of some process models by Laplace transforms: Stirred tank heater

o The model describing the dynamic behaviour of the stirred tank heater was derived and
is given by:
dT Q
Ah  Fi (Ti  T )  ...................................................(15.1)
dt cp

o Heat supplied by steam can be described by the transport rate equation

Q  UAi (Tst  T )..............................................................(15.2)

o So that (1) becomes:

dT UAi
Ah  Fi (Ti  T )  (Tst  T )..........................................(15.3)
dt cp
o Also, if the stirred tank is steady with respect to the liquid level 𝒉 then:
dV
Fi  F and  0 i.e. V  const...........................................(15.4)
dt
 Solution of some process models by Laplace transforms: Stirred tank heater
dT UAi
V  Fi (Ti  T )  (Tst  T )..........................................(15.5)
dt cp
dT UAi UAi
V  FT  FT  T  T .....................................(15.6)
dt
i i i
cp st
cp
dT  UAi  UAi
V    Fi   T  FT  Tst .................................(15.7)
dt  cp  i i
cp

o Divinding throughout by 𝑉 then:

dT F UAi  Fi UAi
    T  T  Tst .................................(15.8)

 V V cp  V c p
i
dt V

o Defining the following parameters:

 Solution of some process models by Laplace transforms: Stirred tank heater

V UAi
 , K= .................................................................(15.8)
Fi V cp
o Then equation 15.8 yields;

dT  1  1
   K  T  Ti  KTst ..........................................(15.9)
dt    
1
o Furthermore, if a = + 𝐾, then;
𝜏

dT 1
 aT  Ti  KTst ..............................................(15.10)
dt 
o At steady state:
1
0  aTs  Ti ,s  KTst ,s ..............................................(15.11)

 Solution of some process models by Laplace transforms: Stirred tank heater

o Subtracting 15.11 from 15.10 yields,

d (T  Ts ) 1
 a (T  Ts )  (Ti  Ti ,s )  K (Tst  Tst ,s )......................(15.12)
dt 
o Defining the deviation variables as follows:

T   T  Ts , Ti  Ti  Ti ,s and Tst  Tst  Tst ,s ........................(15.13)

o 𝑇, 𝑇𝑖 and 𝑇𝑠𝑡 may change with time and are therefore functions of 𝑡𝑖𝑚𝑒.
o Using deviation variables as defined in 15.13 above, equation 15.12 becomes:
dT  1 
 aT   Ti  KTst ..............................................(15.14)
dt 
o Equation 15.14 is a first-order linear differential equation with constant coefficients.
o Laplace transform can be used to obtain the solution of this eq. which gives the dynamic
behaviour of the stirred tank heater.
 Solution of some process models by Laplace transforms: Stirred tank heater

o Assuming that the tank is initially at steady state i.e. 𝑇 ′ 0 = 0. At time 𝑡 > 0, the inlet
temperature 𝑇𝑖 increases by a step of magnitude 10oF and remains at this new
temperature. Thus, 𝑇 ′ 𝑖 t = 10oF. The temperature of the liquid in the tank will start
increasing and it is required to know how this temperature will change with time (dynamic
behaviour).
o Taking the Laplace transform of both sides of equation 15.14 yields;

 dT  

 dt 
 aT  
1
 T   K
i Tst ....................(15.15)
1 
  
sT ( s )  T (0)  aT ( s )  Ti ( s )  KTst ( s )............................(15.16)

o The tank is initially at steady state therefore, 𝑇 ′ (0ሻ = 0.
 Solution of some process models by Laplace transforms: Stirred tank heater

o Furthermore, nothing is said about the change in steam temperature, hence it stays the
same so that 𝑇ത𝑠𝑡
′
(𝑠ሻ = 0. Since, 𝑇 ′ 𝑖 𝑡 = 10oF, then;

T  (t )   T  ( s )  10 .........................................(15.17)
 i  i s
o Thus, equation 15.16 yields;
10
sT ( s )  aT ( s )  .............................................(15.18)
s
10
T ( s )  s  a   ..................................................(15.19)
s
10 1 10  1 1 
T ( s )    .  ......................(15.19)
 s s  a   s s  a 
o The function 𝑇 ′ 𝑡 whose Laplace transform is given by the RHS of 15.19 is the solution
and hence the dynamic behaviour.
 Solution of some process models by Laplace transforms: Stirred tank heater
1 1
o To find the inverse transform of the RHS of equation 15.19 requires breaking .
𝑠 𝑠+𝑎
into partial fractions. Therefore;
1 1 C C2
.  1 .................................(15.20)
s s  a s s  a
o To compute 𝐶1, multiply both sides by 𝑠 and set 𝑠 = 0.
1 1 C1 C2 s
. .s  .s  .................................(15.21)
s s  a s s  a
1 C2 s
 C1  ...........................................(15.22)
o 𝑠 = 0 then, s  a s  a
1 1
 C1  0 or C1  ...........................................(15.23)
a a
 Solution of some process models by Laplace transforms: Stirred tank heater

o To compute 𝐶2, multiply both sides by 𝑠 + 𝑎 and set 𝑠 = −𝑎.

1 1 C1 C2
. ( s  a)  ( s  a)  ( s  a)...........(15.24)
s s  a s s  a
1 C1
 ( s  a)  C2 ....................................................(15.25)
o 𝑠 = −a then, s s
1 1
  0  C2 or C2   ...............................................(15.26)
a a
o Therefore, eq. 15.19 yields,
10 1 10  1 1 1 1 
T ( s )    .  .  ......................(15.27)
 s s  a   a s a s  a 

10  1 1 
T ( s )     ........................................(15.28)
a  s  s  a  
 Solution of some process models by Laplace transforms: Stirred tank heater

o Finally, taking the inverse transform of 15.28 gives the required solution. That is,
10 1 1 
1

T ( s )   1
   ........................................(15.29)
a  s s  a 
10
T (t )  1  e  at .....................................................(15.30)
a
o Thus, to obtain the solution of a differential equation using Laplace transform:
1) Express the differential equation in terms of the deviation variables (if required) and
take the Laplace transform.
o Initial conditions are incorporated at this stage.
2) Solve the resulting algebraic equation in terms of the Laplace transform of the unknown
function.
 Inverse of Laplace transforms by partial fractions
3) Find the time function that has the Laplace transform given by the RHS of the equation
in step 2.
o The third step is the most tedious step. For example given the following Laplace
transform;
 s 2  a1s  b1  (s  c1 )
x ( s)  ...............................(15.31)
s ( s  a2 s  b2 s  c2 )
3 2

o It may not be obvious what function 𝑥(𝑡ሻ has its Laplace transform given by equation
15.31.
o And inversion by partial fractions to find the time domain function 𝑥(𝑡ሻ may not be
obvious.
o Thus, determining the inverse Laplace transform in a crucial step in finding the solution.
o General procedure may be summarised as
 Inverse of Laplace transforms by partial fractions
o Assuming that the resulting Laplace transform has the following form:
Q( s)
x ( s)  ...........................................(15.32)
P( s)
1) Expanding the RHS into partial fractions yields;

Q( s) C C C C
x ( s)   1  2  3  .... n ................(15.33)
P( s ) r1 ( s ) r2 ( s ) r3 ( s ) rn ( s )

2) Where 𝑟1 (𝑠ሻ, 𝑟2 (𝑠ሻ, 𝑟3 (𝑠ሻ are the lower-order polynomials such as first- or second-
order.
3) Find the inverse Laplace transform of every resulting partial fraction, that is,

 C1   C1   C1   Cn 
x(t )  1
 x (s)  1
 
1
 
1
   ....
1
  ....(15.34)
 r1 ( s )   r1 ( s )   r1 ( s)   rn ( s) 
 Inverse of Laplace transforms by partial fractions

o ҧ ሻ is usually a ratio of two polynomials.

𝑥(𝑠
o Its expansion into partial fractions depends on the nature of the roots of the
polynomial 𝑃(𝑠ሻ.
o Three cases can therefore be distinguished:
1) Distinct real roots of the polynomial.
2) Distinct complex roots.
3) Repeated or multiple roots of the polynomial.
1. Distinct real roots
Q( s)
x ( s)  ...........................................(15.32)
P( s)
 Inverse of Laplace transforms by partial fractions: Different real roots

o Suppose the Laplace transform of the function is given by:

s2  s  6
x ( s)  3 ...........................................(15.35)
s  2s  s  2
2

P( s )  s 3  2s 2  s  2...........................................(15.36)
o The roots of the third-order polynomial, eq. 15.36, can be found, for example, by
synthetic division and are as follows:
s  1, s  1 and s  2 or the factors are: s -1, s  1, s - 2....................(15.37)

o Therefore;
s2  s  6 s2  s  6
x ( s)  3  .....................................(15.38)
s  2s  s  2 ( s  1)( s  1)( s  2)
2
 Inverse of Laplace transforms by partial fractions: Different real roots

o Expressing 15.38 into partial fractions:

s2  s  6 C C C
 1  2  3 ............................(15.39)
( s  1)( s  1)( s  2) s  1 s  1 s  2
o Inverting each partial fraction, the solution is then given by:

x(t )  C1et  C2e t  C3e 2t ..................................................(15.40)

o To compute 𝐶1 : multiply both sides of 15.39 by 𝑠 − 1 and set 𝑠 = 1.

s 2
 s  6   s  1
 C1 
C2  s  1 C3  s  1
 ............................(15.41)
( s  1)( s  1)( s  2) s 1 s2
6
And  C1  0  0 or C1  3
2
 Inverse of Laplace transforms by partial fractions: Different real roots

o To compute 𝐶2 : multiply both sides of 15.39 by 𝑠 + 1 and set 𝑠 = −1.

2
C2  
3
o Finally, to compute 𝐶3 : multiply both sides of 15.39 by 𝑠 − 2 and set 𝑠 = 2.
4
C3  
3
o Therefore;
2  t 4 2t
x(t )  3e  e  e ..................................................(15.42)
t

3 3
 Inverse of Laplace transforms by partial fractions: Distinct complex roots

2. Distinct complex roots: Consider the function 𝑥(𝑡ሻ with the Laplace transform:
s 1
x ( s)  2 ...........................................(15.43)
s  2s  5
o The polynomial 𝑃(𝑠ሻ is second-order and has two distinct roots which are not real, but
rather complex conjugates.
b  b 2  4ac
s ...........................................(15.44)
2a
(2)  (2) 2  4(1)(5)
s
2(1)
2  16 2  4 1
s = or s  1  j 2 and 1  j 2
2 2
o That is, complex conjugates.
 Inverse of Laplace transforms by partial fractions: Distinct complex roots

s 1 s 1
x ( s)   ...................(15.45)
s  2 s  5  s  1  j 2    s  1  j 2  
2

o Expanding the RHS into partial fractions;

s 1 C1 C2
 + ...........(15.46)
 s  1  j 2    s  1  j 2    s  1  j 2    s  1  j 2  

o To compute 𝐶1 : Multiply both sides by 𝑠 − 1 + 𝑗2 and set 𝑠 = 1 + 𝑗2 :

 s  1  s  1  j 2  C2  s  1  j 2  
 C1 + ..............................(15.47)
 s  1  j 2    s  1  j 2    s  1  j 2  
 1  j 2   1  C + C2 1  j 2   1  j 2  ..............................(15.48)
1  j 2   1  j 2   1  j 2   1  j 2  
1

 2  j 2   C +0 or C   2  j 2   1  j  j =  j  1  1  j  ......................(15.49)
2
1 1
j4 j4 j2 j 2
 Inverse of Laplace transforms by partial fractions: Distinct complex roots

o To compute 𝐶2 : Multiply both sides of eq. 15.46 by 𝑠 − 1 − 𝑗2 and set ሾ𝑠 =

 s  1  s  1  j 2  C1  s  1  j 2  
  C2 ..............................(15.50)
 s  1  j 2    s  1  j 2    s  1  j 2  
 1  j 2   1  C1 1  j 2   1  j 2   C ..............................(15.51)
1  j 2   1  j 2   1  j 2   1  j 2  
2

C2 
 1  j 2   1
=
2  j2
=
2  j2 1 j
= .........(15.52)
1  j 2   1  j 2   1  j 2   1  j 2    j 4 2

C1 C2
x ( s)  + .....................................(15.54)
 s  1  j 2    s  1  j 2  
 Inverse of Laplace transforms by partial fractions: Distinct complex roots

o Then

 C1 
 
 C2 

1
 x ( s)  1
 +
1
 ...............(15.55)
  s  1  j 2   
    s  1  j 2   
 
x(t )  C1e
1 j 2 t
+C2e
1 j 2 t
...............................................................(15.56)

o Substituting 𝐶1 and 𝐶2 .
1  j 1 j 2t 1  j 1 j 2t
x(t )  e + e .............................................(15.57)
2 2
1  j j 2t t 1  j  j 2t t
x(t )  e .e + e .e ...........................................(15.57)
2 2
et
x(t )  1  j  e j 2t + 1  j  e  j 2t  ......................................(15.57)
2
 Inverse of Laplace transforms by partial fractions: Distinct complex roots

o Using Euler’s identity:

e j  cos  j sin  ..................................................(15.58)
e  j  cos  j sin  ................................................(15.59)
e j 2t  cos 2t  j sin 2t..............................................(15.60)
e  j 2t  cos 2t  j sin 2t.............................................(15.61)

o Substituting 15.60 and 15.61 into 15.57.

et
x(t )  1  j  cos 2t  j sin 2t  + 1  j  cos 2t  j sin 2t   .............(15.62)
2
et  cos 2t  j sin 2t  j cos 2t  sin 2t  
x(t )    ..........................(15.63)
2  +  cos 2t  j sin 2t  j cos 2t  sin 2t  
x(t )  et  cos 2t  sin 2t ...............................................................(15.64)
 Inverse of Laplace transforms by partial fractions: Distinct complex roots

o Finally, using the identity;

a3 sin(b   )  a1 cos b  a2 sin b.....................................(15.65)
 a 
with a3  a12  a22 and   tan 1   1 
 a3 
Therefore, cos 2t  sin 2t  2 sin(2t   )
 x(t )  et 2 sin(2t   )

o Thus, complex conjugate roots will give rise to a periodic function i.e., sinusoidal
function.
 Inverse of Laplace transforms by partial fractions: Repeated roots

o Expansion into partial fractions and computation of the coefficients in this case is
different when the polynomial 𝑃(𝑠ሻ contains multiple roots.
o As an example, the Laplace transform of the function 𝑥(𝑡ሻ given below has repeated or
multiple roots:
1
x ( s)  ..............................................(15.66)
( s  1) ( s  2)
3

o Expansion into partial fractions is as follows:

1 C1 C2 C3 C4
= + +  ........................(15.67)
( s  1) ( s  2) s  1 ( s  1) ( s  1)
3 2 3
s2
o The polynomial 𝑃(𝑠ሻ has three equal roots:

s1  s2  s3  1 and s4  2...........................................(15.68)
 Inverse of Laplace transforms by partial fractions: Repeated roots

o The inverse Laplace transform of eq. 15.67 is simply:

x(t )=C1e  t +C2te  t +C3t 2e  t  C4e 2t ........................(15.69)

o Compute 𝐶4 : Coefficient 𝐶4 corresponds to a distinct root and can be computed as

before.
o Multiply both sides of 15.67 by 𝑠 + 2 and set 𝑠 = −2.
1 C1 C2 C3
=  s  2 + 2 
s  2 + 3
s  2   C4 ........................(15.70)
( s  1) s  1
3
( s  1) ( s  1)
1
=0+0+0  C4 or C4  1
(2  1) 3

o Compute 𝐶3 : Applying the method as before i.e. multiply 15.67 by 𝑠 + 1 3 and set 𝑠 =
− 1.
 Inverse of Laplace transforms by partial fractions: Repeated roots
1 C4
=C1 ( s  1) +C2 ( s  1)+C3 
2
( s  1)3.............................(15.71)
( s  2) s2
1 C4
=C1 (1  1) +C2 (1  1)+C3 
2
( 1  1)3.............................(15.72)
(1  2) 1  2
C3  1
o Compute 𝐶2 : The coefficient 𝐶2 cannot be computed as before since multiplying 15.67
1
by (𝑠 + 1ሻ2 and setting 𝑠 = −1 will result in which is undefined (check).
0

o Differentiate both sides of 15.71 with respect to 𝑠 and set 𝑠 = −1.

1  ( s  1)3 3( s  1) 2 
 =2C1 ( s  1)+C2  C4     .............................(15.73)
( s  2) 2
 ( s  2)
2
s2 
1  (1  1)3 3(1  1) 2 
 =2C1 (1  1)+C2  C4     or C2  1
(1  2) 2
 (1  2)
2
1  2 
 Inverse of Laplace transforms by partial fractions: Repeated roots

o Compute 𝐶1 : Differentiate 15.73 with respect to 𝑠 and set 𝑠 = −1.

2  ( s  1)3 3( s  1) 2 3( s  1) 2 6( s  1) 
=2C1  C4  2     ...............(15.74)
( s  2) 3
 ( s  2)
3
( s  2) 2
( s  2) 2
s2 
2  (1  1)3 3(1  1) 2 3(1  1) 2 6(1  1) 
=2C1  C4  2     or C1  1
(1  2) 3
 (1  2)
3
(1  2) 2
(1  2) 2
1  2 
o Therefore:
x(t )=e  t  te  t  t 2e  t  e 2t ..................................(15.75)
 Second-order differential equation using Laplace transform

o Consider the following general form of a second-order differential equation.

d 2x dx
a2 2  a1  a0 x  f (t )........................................15.76
dt dt
o Where the function 𝑥(𝑡ሻ is considered to be in deviation form and has an initial
condition;
dx
x(0)   0  x(0)........................................15.77
dt t 0

o Taking the Laplace transform of both sides of 15.76 yields;

 d 2x   dx 
a2  dt 2  1  dt   a0
 a  x    f (t ).........................15.78
   

a2  s 2 x ( s )  sx(0)  x(0)   a1  sx ( s )  x(0)   a0  x ( s )   f ( s )..............15.79

 Second-order differential equation using Laplace transform

o Since dx
x(0)   0  x(0)........................................15.77
dt t 0

o Then; a2  s 2 x ( s )   a1  sx ( s )   a0  x ( s )   f ( s )..................15.80

o Assuming that 𝑓(𝑡ሻ is a step function of magnitude 1, i.e. unit step function. Then;
1
f (t )  1 and f ( s )  .......................................15.81
s
o Equation 15.80 becomes;
1
 
x ( s )  a2 s  a1s  a0   .............................15.82
2

s
1
x ( s)  .............................15.83
s  a2 s  a1s  a0 
2
 Second-order differential equation using Laplace transform

o The polynomial 𝑃∗ 𝑠 = 𝑎2 𝑠 2 + 𝑎1 𝑠 + 𝑎0 in equation 15.83 is called the

characteristic polynomial and is second-order.
o To invert 15.83, the roots of the polynomial 𝑃∗ 𝑠 have to be determined.
o Depending on the values of the coefficients 𝑎2 , 𝑎1 𝑎𝑛𝑑 𝑎0 , three cases can be
distinguished.
a1  a12  4a2 a0
s1,2  .........................................15.84
2a2
o Case 1: when 𝑎12 − 4𝑎2 𝑎0 > 0.
o Results in two distinct roots and the computation of the partial fractions coefficients can
be performed as in the case for distinct roots already discussed.
 Second-order differential equation using Laplace transform

o Case II: when 𝑎12 − 4𝑎2 𝑎0 = 0. Results in two equal or repeated roots.
o Computation of the partial fractions coefficients can be performed as in the case for
repeated or multiple roots already discussed.
o Case III: when 𝑎12 − 4𝑎2 𝑎0 < 0. Results in two distinct complex conjugate roots.
o Computation of the coefficients in the resulting partial fractions can be performed as in
the case for distinct complex conjugate roots previously described.