UCH2504 Process Instrumentation, Dynamics

and Control
Laplace Transform for Process Control

CTRL
Process Dynamics

A chemical process is a dynamical system, whose behavior changes

over time. Control systems are needed to handle such changes in
the process. Thus, it is important to understand the process
dynamics when a control system is designed.

Mathematically, the process dynamics can be described by

differential equations. Unsteady-state (or transient) process
behavior corresponds to a situation, where (at least some) time
derivatives of the differential equations are nonzero.

CTRL
Transient Nature of Chemical Processes

Transient operation occurs during important situations such as

start-ups and shutdowns, unusual process disturbances, and
planned transitions from one product grade to another.

Even at normal operation, a process does not operate at a steady

state (with all time derivatives of the differential equations exactly
zero) because there are always variations in external variables, such
as feed composition or cooling medium temperature.

Thus, knowledge of steady-state (or static) process properties,

learned in previous courses (such as thermodynamics, fluid
mechanics, heat transfer, mass transfer, reaction engineering), is
not sufficient for control design.

Dr. M. Subramanian CTRL

Laplace Transform

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

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.

The L operator transforms a time domain function f (t) into an ‘s’

domain function, F(s).

s is a complex variable: s = x + iy.

Dr. M. Subramanian CTRL

Laplace Transform of Common Functions
Function f (t) F (s)
a
Constant a
s
a
Ramp at
s2
n!
tn
sn+1
1
Exponential e−at
s+a

1
eat
s− a
n!
tne−at
(s + a)n+1
Dr. M. Subramanian CTRL
Laplace Transform of Common Functions (contd..)
Function f (t) F (s)

ω
Sinusoidal sin(ωt)
s2 + ω2
s
cos(ωt)
s2 + ω2
ω
Sin with exponential e−at sin(ωt)
(s + a)2 + ω2
s+a
e−at cos(ωt)
(s + a)2 + ω2

ω
Hyperbolic sinh(ωt)
s2 − ω2
s
cosh(ωt)
s2 − ω2
Dr. M. Subramanian CTRL
Laplace Transform of Common Functions (contd..)

Dr. M. Subramanian CTRL

Some Rules

Dr. M. Subramanian CTRL

Some Rules (contd..)

First shifting rule:

L[eat f (t)] = L[f (t)] s→(s−a)

Example

Dr. M. Subramanian CTRL

Laplace Transform of Derivatives
Laplace Transform of Derivatives (contd..)

Dr. M. Subramanian CTRL

Laplace Transform of Derivatives (contd..)

Dr. M. Subramanian CTRL

Laplace Transform of Integrals

Dr. M. Subramanian CTRL

Initial and Final Value Theorems

Initial Value Theorem

lim f (t) = slim [sF(s)]

t→0 →∞

It helps to determine the value of the time function f (t) at

t = 0 without finding the inverse of F(s).
Final Value Theorem

lim f (t) = lim [sF(s)]

t →∞
s→0

It helps to determine the steady-state value of the system

response.

Dr. M. Subramanian CTRL

Exercise Problems

dy
1
Solve by using Laplace transform: = 5 − 2t; y(0) = 1.
dt
(Ans: y = 1 + 5t − t 2 )
2
Solve by Laplace transform method, the initial value problem
y '' = 10; y(0) = y '(0) = 0. (Ans: y = 5t2)
Take Laplace transform on both sides Take Inverse Laplace transform

Solve for Y(s)

2. Solve by Laplace transform method, the initial value problem
y ' ' = 10; y(0) = y '(0) = 0.
Take Laplace transform on both sides

Inverse Laplace transform

UCH2504 Process Instrumentation, Dynamics
and Control
First Order Systems

CTRL
Liquid Level Variation for Changes in In-flow
q(t)

h(t)

R
qo (t)

CTRL
Liquid Level Variation for Changes in In-flow (contd..)

CTRL
Liquid Level Variation for Changes in In-flow (contd..)

CTRL
Liquid Level Variation for Changes in In-flow (contd..)

CTRL
Liquid Level Variation for Changes in In-flow (contd..)

Using the deviation variables in Eqn.(3), we get

H dH
Q= +A (4)
R dt
Taking Laplace transform for the above, we get
H(s)
Q(s) = + AsH(s) (5)
R
Note: L[dH/dt] is simply sH(s), because H(0) = 0. Rewriting the
Eqn.(5) as,
H(s) R
= (6)
Q(s) τs + 1
where τ = AR = time constant of the system.
H(s)
The term = G(s) is called as the transfer function of the
Q(s)
system.

CTRL
Liquid Level Variation for Changes in In-flow (contd..)

CTRL
Liquid Level Variation for Changes in In-flow (contd..)

From the relation between qo and h we can write,

Qo =
H
=⇒ Qo(s) 1
= (7)
R H(s) R

From Eqns.(6) and (7) we get,

Qo(s) 1
=
Q(s) τs + 1

CTRL
Response of Thermometer

CTRL
 Response of Thermometer
Assumptions:
1. All the resistance to heat transfer resides in
the film surrounding the bulb (i.e., the
resistance offered by the glass and mercury is
neglected).
2. All the thermal capacity is in the mercury.
Furthermore, at any instant the mercury
assumes a uniform temperature throughout.
3. The glass wall containing the mercury does
not expand or contract during the transient
response.
4. Thermometer is initially at steady state
(i.e., before time zero, there is no change in
temperature with time. At time zero, the
thermometer will be subjected to some change
in the surrounding temperature x(t).
CTRL
Response of Thermometer

𝑇∞ = 𝑓𝑙𝑢𝑖𝑑 𝑡𝑒𝑚𝑝
T= thermometer reading
𝑊 𝐽
*𝑚2 *K = kg* ∗ 𝐾/𝑠
𝑚2 𝐾 𝐾𝑔∗𝐾

CTRL
First Order System (Generalization)

CTRL
First Order System

CTRL
First Order System (contd..)

If Y (t) = y(t) − ys and X(t) = x(t) − xs are in terms of deviation

variables around a steady state, then the initial conditions are:

Y (0) = y(0) − ys = ys − ys = 0 and X(0) = x(0) − xs = 0

Using the above conditions, and taking the Laplace transform for
Eqn.(1), we get
Y (s) Kp
G(s) = =
X(s) τps + 1
Because of the usage of deviation variables, the Laplace transform
of the differential equation results in an equation that is free of
initial conditions, because the initial values of X and Y are zero.

CTRL
Transfer Function

Y (s) Kp Laplace transform of output deviation

= = = G(s)
X(s) τps + 1 Laplace transform of input deviation
The above ratio is called the transfer function, G(s), of the
system. In examining physical systems, we usually attempt to
obtain a transfer function.

CTRL
Transfer Function (contd..)

Procedure for obtaining the transfer function for a process:

11 1 Write the appropriate balance equations (usually mass or

energy balances for a chemical process).
St 2. Linearize terms if necessary.
St 3. Write the balance equations in deviation variable form.
St 4 Take Laplace transform for the linear balance equations.
St 5. Rearrange the resulting transformed equation into the transfer
function form (i.e., the output divided by the input).

CTRL
Idealized Input Functions

(a) step input (c) sinusoidal input

(b) ramp input (d) impulse input

CTRL
Idealized Input Functions (contd..)

CTRL
First Order System - Response to Step Input
The transfer function of first order system is given by
Y (s) Kp
G(s) = =
X(s) τps + 1
Let us examine how it responds to a unit step change in input,
x(t).
1
For unit step input, X(s) = . Therefore,
s
Kp Kp Kp τp
Y (s) = = −
s(τps + 1) s τps + 1
Taking inverse Laplace transform, we get
y(t) = Kp (1 − e −t/τ p )
For step input of magnitude A,

y(t) = AKp (1 − e −t/τ p )

CTRL
Taking Inverse Laplace
First Order System - Response to Step Input (contd..)

If the initial rate of change of Y (t) were maintained, the

response would be complete in one time constant.
The value of Y (t) reaches 63.2% of its ultimate value when
the time elapsed is equal to one time constant τp.
CTRL
Solved Problems

Example 1: Tank Dynamics for Step Change in In-flow

A tank of volume 0.25 m3 and height 1 m has water flowing in at
0.05 m3/min. The outlet flow rate is governed by the relation
Fout = 0.1h where h is the height of the water in the tank in m
and Fout is the outlet flow rate in m3/min. The inlet flow rate
changes suddenly from its nominal value of 0.05 m3/min to 0.15
m3/min and remains there. The time (in minutes) at which the
tank will begin to overflow is given by (G-2008-62)

(a) 0.28 (b) 1.01 (c) 1.73 (d) ∞

CTRL
Solved Problems (contd..)

Solution:
From balance on volumetric flow rate,
d(Ah) dh
Fi − Fo = =A
dt dt
i.e.,
dh
A + Fo = Fi
dt
Given: Fo = Fout = 0.1h m3/min. And,
A = Vtotal/htotal = 0.25/1 = 0.25 m2. Therefore, the above
equation becomes
dh
0.25 + 0.1h = Fi (1)
dt

CTRL
Solved Problems (contd..)

At initial steady state, dh/dt = 0 and, Fi = Fo = 0.1ho .

Given: Fi = 0.05 m3/min; Therefore, ho = 0.05/0.1 = 0.5 m.
Rewriting the Eqn.(1) as below:

dh
2.5 + h = 10 Fi (2)
dt
At the initial steady state, the above equation becomes,

0 + ho = 10 × Fio

i.e.,
0 + 0.5 = 10 × 0.05 (3)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
Solved Problems (contd..)

CTRL
CTRL
CTRL
Response of First Order System to Step Input

CTRL
Impulse Input to First Order System

CTRL
Impulse Input to First Order System (contd..)

CTRL
Comparison of Responses of Unit Step and Impulse Inputs

CTRL
Tutorial-1

CTRL
Tutorial-1

CTRL
Tutorial-1

CTRL
Tutorial-1

CTRL
Tutorial-2

CTRL
Tutorial-2

CTRL
Tutorial-2

CTRL
Stirred Tank Heater

ṁ,Ti

m,T

ṁ,T

q
heating coil

Assume:
ṁ = mass flow in = mass flow out = constant.
m = mass of tank contents = constant
To find: the variation of T with t for changes in q and/or Ti .
CTRL
Stirred Tank Heater (contd..)

CTRL
Stirred Tank Heater (contd..)

CTRL
Stirred Tank Heater (contd..)

CTRL
Response of First Order System to Ramp Input

CTRL
Response of First Order System to Ramp Input (contd..)

CTRL
Liquid Level System with Constant Outflow

Consider the case where there is a pump in the outflow line. Here
the outflow doesn’t depend on the head of liquid available, and it
is constant.

CTRL
Liquid Level System with Constant Outflow (contd..)

From mass balance,

dh
q − q0 = A (1)
dt
At steady state,
qs − qos = 0 (2)
Subtracting Eqn.(2) from (1), and using deviation variables, we get
dH
Q=A (3)
dt
where Q = q − qs; and, H = h − hs . Taking Laplace transform for
Eqn.(3), we get
H(s) 1
=
Q(s) As

CTRL
Liquid Level System with Constant Outflow (contd..)

The integral of a step is a ramp

CTRL
Linearization

• All the examples of physical systems discussed so far ,

including the liquid-level system have been linear. Actually,
most physical systems of practical importance are nonlinear.

• Characterization of a dynamic system by a transfer function

can be done only for linear systems

• For dynamic analysis, non linear systems can be

approximated to linear ones. A very important technique
for such approximation is illustrated by the following
discussion of the liquid-level system

CTRL
Linearization

CTRL
Linearization

CTRL
Linearization

CTRL
UCH 2504 Process Dynamics and
Control
First Order Systems in Series

CTRL
First Order Systems in Series

CTRL
(a) Non-interacting Systems in Series
Tank-1:
Material balance:
dh1
q − q1 = A1 (1)
dt
Relation between q1 and h1:
h1
q1 = (2)
R1
Tank-2:
Material balance:
dh2
q1 − q2 = A2 (3)
dt
Relation between q2 and h2:
h2
q2 = (4)
R2
Writing the above equations in deviation variables
{Q = q −q s ;Q1 = q1 −q 1s H1 = h1 −h 1s ; H2 = h2 −h 2s }, taking
Laplace transforms, and using A1R1 = τ1; A2R2 = τ2, we get
CTRL
 Taking Laplace both sides
Assume zero initial condition

Differentiate both sides

Substitute this value in 1st eq

Multiply both sides by A1

Multiplying the above 2 eqs
(a) Non-interacting Systems in Series (contd..)

H1(s) R1
= (5)
Q(s) τ1s + 1
Q1(s) 1
= (6)
Q(s) τ1s + 1
H2(s) R2
= (7)
Q1(s) τ2s + 1
And,
H2(s) 1 R2
= (8)
Q(s) τ1s + 1 τ2s + 1
Q2(s) 1 1
= (9)
Q(s) τ1s + 1 τ2s + 1
Eqn.(9) can be generalized for multiple (non-interacting) systems
in series as
QN (s) Kp1 Kp2 KpN
= ··· (10)
Q(s) τ1s + 1 τ2s + 1 τN s + 1
CTRL
(a) Non-interacting Systems in Series (contd..)

CTRL
(a) Non-interacting Systems in Series (contd..)

CTRL
(b) Interacting Systems in Series

CTRL
Tank2

Solve the above equation for H2(s)

Substitute the above eq in Q(s)

Substitute H2(s) in tank 2

 From this equation

Substitute the above equation in

Let
(b) Interacting Systems in Series (contd..)

CTRL
(b) Interacting Systems in Series (contd..)

CTRL
Comparison between Non-interacting and Interacting
Systems

CTRL
Comparison (contd..)

Here, τ1 = τ2, and A1 = A2.

Interacting system is more sluggish than the noninteracting system.
CTRL
 From eq 3
Substitute H1(s) in eq 1
Substitute H2(s) in eq 2

Cancel R1

Simplifying

Solve for H1(s)

 (t) Y(t)q
q

V, y(t), V
UCH 2504 Process Dynamics and
Control
Second Order Systems

CTRL
Second Order System

CTRL
Damping Factor

CTRL
Response of Second-order system to Step Input

1.5

y(t) 1
Kp

0.5

0
0 5 10 15 20 25 30
t/τ
CTRL
Response of Second Order System to Unit Step Input
Overdamped System

CTRL
Overdamped System

The system response resembles a little the response of a

first-order system to a unit step input. But when compared to
a first-order response we notice that the system initially delays
to respond and then its response is rather sluggish.
Overdamped responses are the responses of multi-capacity
processes, which result from the combination of first order
systems in series.

CTRL
Response of Second Order System to Unit Step Input
Critically Damped System

CTRL
Response of Second Order System to Unit Step Input
Underdamped System

CTRL
Response of Second Order System to Unit Step Input

If ζ = 0, then such a second-order system is marginally stable

in that the response is of constant amplitude in time. This is
the undamped case.
If ζ < 0, then such a second-order system is unstable and the
response grows in time without bound.

CTRL
Characteristics of an Underdamped Response

CTRL
Characteristics of an Underdamped Response (contd..)

CTRL
Characteristics of an Underdamped Response (contd..)

CTRL
Problem
Problem
Dynamics of U-tube Manometer
•Consider a simple manometer as shown in Fig. . The pressure on both legs of the
manometer is initially the same. The length of the fluid column in the manometer
is L.
•At time t =0, a pressure difference is imposed across the legs of the manometer. Assuming
the resulting flow in the manometer to be laminar and the steady-state friction
law for drag force in laminar flow to apply at each instant, we will determine the
transfer function between the applied pressure difference ΔP and the manometer
reading h.
•If we perform a momentum balance on the fluid in the manometer, we arrive at the
following terms:
( Sum of forces causing fluid to move) =(R ate of change of momentum of fluid)

CTRL
Dynamics of U-tube Manometer

The term for the skin friction at the wall is obtained from the Hagen-Poiseuille relationship for laminar flow. Note
that V is the average velocity of the fluid in the tube, which is also the velocity of the interface, which is equal to 1/2
dh/dt
CTRL
The rate of change of momentum of the fluid may be expressed as

Rearranging the above equation

Note that as with first-order systems, standard form has a coefficient of 1 on the
dependent variable term, h in this case. Second-order systems are described by a second order
differential equation.
Step Response:

If the forcing function is a unit-step function, we have

…. (1)

In terms of the manometer , this is equivalent to suddenly applying a

pressure difference [such that X ( t ) =ΔP / ρg =1] across the legs of the
manometer at time t = 0.
• Combining Eq. (1) with the transfer function of gives
…. (2)

• The quadratic term in this equation may be factored into two linear
terms that contain the roots
τ2s2 + 2ζτs + 1 = 0
2 1
s2 + s+ 2 = 0
 
 2 4 2
4
−  −
   =−  1
2 2
s=  2 −1
2  

s1 = (−  +
1

 2 −1 ) ....(3)

s = (−  − −1 )
1
2 ....(4)

1

Equation (2) can now be written

….. (5)

The response of the system Y(t) can be found by inverting Eq. (5).
The roots s1 and s2 will be real or complex depending on the
parameter ζ.
• The problem may be divided into three cases shown in
Table.
The damping of a system can be described as being one of the following:
Overdamped: The system returns to equilibrium without oscillating.
Critically damped: The system returns to equilibrium as quickly as possible
without oscillating.
Underdamped: The system oscillates (at reduced frequency compared to
the undamped case) with the amplitude gradually decreasing to zero.
Undamped: The system oscillates at its natural resonant frequency (ωo).
CASE I STEP RESPONSE FOR ζ > 1: (Over-damped system). For
this case, the inversion of Eq.(5) gives
 
Y (t ) =
1 1
L−1  
2  s (s − s 1 )(s − s )
2  …(6)

1 A B C
= + +
s(s − s1 )(s− s2 ) s s − s1 s −s2
A(s − s1 )(s− s2 )+ Bs(s − s2 )+ Cs(s − s1 )= 1

1 1 2
Put s = 0  A =
s1s2
=
1
2
(−  + )(
 −1 −  −  −1
2 2
) =
 −  2 +1
2
= 2

1  2
Put s = s1  B = =
(
s1 (s1 − s2 ) −  +  2 − 1 2  2 − 1 )( )
1 2
Put s = s2  C =
s2 (s2 − s1 )
=−
(
−  −  2 −1 2  2 −1 )( )
 • From Eq(6)

 1 2 2 1 
Y (t ) =
1 1
2
L−1  2 +
2
(
2
−
2 2
)( 
 s −  +  − 1 2  − 1 s − s1 −  −  − 1 2  − 1 s − s2  ) ( )( )
1   2 2 2 
Y (t ) = es t  
 +
( es t −
)( ) ( )( )
1 2

 2   −  +  −1 2  −1
2 2
−  −  −1 2  −1
2 2
 

−t  2 −1
−t −
 2 −1
  t   t
e e e e
Y (t ) = 1 +
(−  + )(
 2 −1 2  2 −1 ) − (−  −  − 1)(2  − 1) 2 2

( − 1 )e − (−  +  − 1 )e
−t  2 −1
 −
 2 −1

 − − 
 2  t 2  t
e 
Y (t ) = 1 +
 2 −1  (
2  2 −  2 +1 ) 
 
−t
e   
  2 −1
−
 2 −1
  2 −1   2 −1
−
 2 −1

Y (t ) = 1 + −  e −e − e −e 
 t  t  t  t

 − 1  2 
2
 2  
 −t

(
 −1 t −  2 −1 )  2 −1 t
2
e
Y(t) =1+ − sinh cosh
 −1
2   

….(7)

where the hyperbolic functions are defined as

• The response is non-oscillatory and becomes more “sluggish” as ζ

increases. This is known as an over-damped response

• The nature of roots for an over-damped system is real

• CASE II STEP RESPONSE FOR ζ= 1: (Critically Damped system) Put ζ=1
in Eq(7),
 sinh  2 − 1. t 
Y (t ) = 1 − e
−t
 cosh 0 + 
 lim
 →1  2 −1 
 
 sinh x. t 
Y (t ) = 1 − e
−t
 cosh 0 + lim 
 x →0 x 
 
By L- Hospital’s rule,

t t
sinh x cosh x
 =  t =t
lim
x →0 x lim
x →0 1  

y (t ) = 1 − e − t / 1 + t (  )
The response is nonoscillatory and the nature of the roots will be real,
and equal.
• CASE III STEP RESPONSE FOR ζ< 1: (Under Damped system)
 2 −1 = (− 1)(1 −  2 )= j 1 − 2
….(8)

cosh j = cos  ; sinh j = j sin  ….(9)

Use Eqns (8) & (9) in (7)

−t
 1−  2
 1−  2 
Y (t ) = 1 − e 
cosh j t+ sinh j t
  j 1−  2  
−t
e   1−  2
1−  2 
Y (t ) = 1 −  j 1−  2
cos t + j sin t
j 1−  2
   
−t
e   1−  2
1−  2 
Y (t ) = 1 −  1−  2
cos t +  sin t ....(10)
1−  2
   
 • The square bracket in Eq (10) is of the form p Cos A + q Sin A .

• The solution of this form is r sin (A+θ), Where

r = p 2
+ q 2

p
 = tan −1
q

Eq (10) can be written as

−t
e   1−  2
1−  2 
 Y (t ) = 1 − sin  t + tan −1  …..(11
1−  2    
  )
Where
r = 1−  2 +  2 =1
1−  2
 = tan−1

The response is either under damped or oscillatory and the nature of the roots
will be complex.
• Impulse
Response:
If a unit impulse δ(t) is applied to the second-order system, then the transform of
the response is

…(17)

• As in the case of the step input, the nature of the response to a unit impulse
will depend on whether the roots of the denominator of Eq. (17) are real or
complex.

CASE I IMPULSE RESPONSE FOR ζ < 1. The inversion of Eq. (17) for ζ < 1
yields the result

….(18)

• A simple way to obtain Eq. (18) from the step response of Eq. (11) is to take

the derivative of Eq. ( 11) with respect to t.

 −t  1−  2  1−  2   1 −  2 
 +  cos 1 − 
dyStep  2
e
Y (t ) = =− − 1 −  2 sin t t 
dt 1 −  2         
   
Im pulse

 e −     
t
 1−  2 1−  2
+  1 −  2 cos t +  sin t  −  − 
    1 −  2    
−t
dy e  1−  2   2 
= sin t 1 −  2 +
dt    1 −  2 

−t
e  1−  2
Y (t ) = sin t ...(19)
Im pulse  1−  2 
CASE II IMPULSE RESPONSE FOR ζ = 1. For the critically damped case, the
response is given by
Put ζ = 1 in Eq (19), Y(t) = 0/0,
Use L-Hospital’s rule,

xt
−t sin

Y (t ) =
e 
 lim x →0 x
Where x = 1 −  2

xt  t 
−t cos  
e 
  
Y (t ) = lim
 x →0 1
−t

e t
Y (t ) =  
  

−t

te
Y (t ) = ...( 20)
 2
 • CASE III IMPULSE RESPONSE FOR ζ > 1.

(−1)( )
Substitute 2
1− =
2 2
−1 = j  −1

sinh j = j sinh
in Eq. (19)

−t
e   2 −1
Y (t ) = j sinh t
 . j  −1 2 
−t
e   2 −1
Y (t ) = sinh t ...(21)
  −1 2 

The impulse-response show the same general behavior as the step- response.
However, the impulse response always returns to zero.
Sinusoidal Response
or Frequency
response
• If :the forcing function applied to the second-order system is sinusoidal, X(t)
= A sin ωt then it follows
…(22)

• The inversion of Eq. (22) may be accomplished by first factoring the

two quadratic terms to give

…(23
Here s1 and s2 are the roots of the denominator of the transfer function and
)
are given by Eqs. (3) and (4). For the case of an underdamped system (ζ < l),
the roots of the denominator of Eq. (23) are a pair of pure imaginary roots
(+jω, - jω) contributed by the forcing function and a pair of complex roots
 A  C1s + C2 C3 C 
Y (s) = + + 4
 2  s 2 +  2 s − s1 s − s2 
  2 −1   2 −1
s1 = − + ; s2 = − −
   
 2 2

1
( ) ( )
(C1s + C2 ) s + s + 2  + C3 s 2 +  2 (s − s2 ) + C4 s 2 +  2 (s − s1 ) = 1
 

 2 1 
put s = j  (C1 j + C2 ) −  2 + j  + 2  =1
   
2
(C1 j + C2 ) =
(1 −  
2 2
) + j 2 ...( 24)

2
put s = − j  (− C1 j + C2 ) =
(1 −   2 ) − j 2
2
...( 25)

 − j 4 
( 24) − ( 25)  2C1 j =  2  
 1 −  
2
(
2 2
)
+ ( j 2 ) 
2
 − 2
C1 =  2

(1 −  2
 2
)
2
+ ( j 2 ) 2

( 24) + ( 25)  2C 2 =  2


(
2 1 −  2 2 ) 

(
 1− 
2 2 2
)
+ ( j 2 ) 2 

 (1 −   )
2 2 2
C2 =
(1 −   2 )2 + ( j 2 ) 2
2

put s = s1  C3 (s1 +  2 )(s1 − s2 ) = 1

2

1
C3 =
(s1 − s2 )(s12 +2 )
put s = s2  C3 s2 +  2 ( 2
)(s 2 − s1 ) = 1
1
C3 =
(
(s2 − s1 ) s2 2 +  2 )
   2 (2 ) −1  s   2 1 −  2 2 (
 1 
L−1  2
)
− L   + 2 
A
Y (t ) = 2 
(
1 −   + (2 )
2 2 2 2
)  s 2
+  2
 1 −   + (2 )
2 2 2
( 2
)
 s +  
 1  1  1  1  
+ L−1   + L−1   
 (s1 − s2 ) s1 + 
2 2
( )
 s − s1  (s2 − s1 ) s2 + 
2 2
(
 s − s2   )
 (2 ) cos t +
(
1−  2 2
)
sin t

− 
 1 (
−  
2 2 2
) + (2 )2
1 − ( 
2 2 2
)
+ (2 )
2

Y (t ) = A  
 2  
−  2 −1   2  
−  2 −1 t
   +  t
  −   
+ +
   

 ( s1 − s 2 ) (
s 1
2
+  2
e
) ( s 2 − s1 ) s 2
2
+  2
( e
) 


Y (t ) = A−
(2 ) cos t +
1 −  2 2 
sin t 
( )
(
 1 −   + (2 )
2 2 2 2
) 1 −   + (2 )
2 2 2 2
(  )
   2 −1  t   2 −1 
  
   t
A −t
 e  

e  
+ e   2 − 2 
 (s1 − s2 )
2
(
 s1 + 
2
) ( s2 +   2
)
 
When t →

Y (t ) = A−
(2 ) cos t +
( )
1 −  2 2 
sin t 
(
 1 −  
2
)2 2
+ (2 )
2
( 2
)
1 −   2 + (2 )
2 2


Y (t ) =
A
( )
 1 −  2 2 + (2 )2 sin (t +  )
2

( )
1 −  2 2 + (2 ) 
2 2 
Where p = −2
q = 1 −  2 2
 2 
 = tan −1  − 2 
 1−  
2
• The ratio of the output amplitude to the input amplitude is

This may be greater or less than 1, depending on ζ and τω. This is in

direct contrast to the sinusoidal response of the first order system,
where the ratio of the output amplitude to the input amplitude is
always less than 1.
2. The output lags the input by phase angle |Φ|. The phase angle
approaches 180˚ asymptotically as ω increases. The phase lag of the
first- order system, on the other hand, can never exceed 90˚.
Transportation Lag
Transportation Lag
Transportation Lag
Transportation Lag
Transportation Lag
Transportation Lag
Transportation Lag
 Self Assessment Questions

• Define critically damped system and write the transfer function of the
same
• Define under damped system and over damped system
• Define overshoot
• Define rise time
• Define settling time
• Define decay ratio
• Define response time
• Define period of oscillation
 UNIT II
OPEN LOOP SYSTEMS

Laplace Transform
Laplace Transform enables one to get a very simple and elegant method of solving linear differential
equation by transforming them into algebraic equations. It is well known that chemical processes are
mathematically represented through a set of differential equations involving derivatives of process
states. Analytical solution of such mathematical models in time domain is not only difficult but
sometimes impossible without taking the help of numerical techniques. Laplace Transform comes as a
good aid in this situation.
For this reason, Laplace Transform has been included in the text of this ―Process Control‖ course
material though it is purely a mathematical subject.

Definition of Laplace Transform

Consider a function f(t). The Laplace transform of the function is represented by f(s) and defined by the
following expression:
(1)

Hence, the Laplace Transform is a transformation of a function from the t -domain (time domain) to s -
domain (Laplace domain) where both t and s are independent variables.

Laplace transform of a few basic functions

The Fig. 1 shows a few basic functions which are frequently used in process control applications

(a) Step function (b) Ramp function

28 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(c) Pulse function (d) Impulse function

Fig. 1: Few basic functions which are frequently used in process control applications

Step function : See Fig. 1(a) for the schematic of a step function

Hence,

(2)
(3)

Ramp function : See Fig. 1(b) for the schematic of a ramp function f(t)=at for t>0 where a is a constant
(4)

Hence,
(5)

Exponential function : for t>0 where a is a constant

(6)

Hence,
(7)

Sinusoidal function :

(8)

Hence,

29 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(9)

Delayed function : , i.e .f(t) is delayed by seconds

(10)

Now, let us take , hence . At and at .

Thus,
(11)1
1

Hence,
(12)

Pulse function : See 1(c) for the schematic of an unit pulse function. The area under the pulse is 1. The

duration of pulse is T and hence it achieves maximum intensity of . Thus the function is defined

by
It can also be defined as the ―addition‖ of two step functions which are equal but with opposite intensity,
however, the second function is delayed by T .

Hence, it is evident that is equal to in intensity however it is delayed by time

Thus, . Since is a step function of intensity , the following

expression will hold.

Hence,
(14)

Impulse function : See Fig.1(d) for the schematic of an unit impulse function. This is analogous to a
pulse function whose duration is shrinked to zero without losing the strength. Hence the area under the
impulse remains 1. The function can be expressed as the following:
(15)

As the duration of the impulse tends to zero, its maximum intensity ideally tends to . Mathematically
it is termed as Dirac Delta function and is represented as . The following relation holds for unit
impulse:
(16)

Thus the Laplace transform of the impulse function can be derived as the following:

30 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(17)

L'Hospital's rule has been applied in the above derivation. Hence,

(18)

The following table presents the Laplace transforms of various functions.

Laplace transforms of various functions

Function in time domain Laplace Transform
Unit impulse 1

Unit pulse of duration T

Unit step

Ramp : f(t)=at

Laplace Transform of derivatives

The Laplace transform of derivative of a function f(t) is derived in the following manner:
(19)

31 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

Where f(0) is the value of the function at t=0. Similarly it can be proved that
(20)

Where f'(0) is the value of the derivative of the function at t=0. In general, it can be proved that
(21)

Where are the initial conditions of the respective-order

derivatives of the function.

Laplace Transform of Integrals

The Laplace transform of integral of a function f(t) is derived in the following manner:
(22)

Integrate by parts by considering the following: . Then,

and . Hence,
(23
)

Hence,
(24)

Final value theorem

The final value theorem allows one to compute the value that a function approaches as when its
Laplace transform is known.
(25)

Initial value theorem

The Initial value theorem allows one to compute the value that a function approaches as when its
Laplace transform is known.
(26)

Solution of linear ODEs using Laplace Transform

Following example illustrates the method of solving ODEs using Laplace Transform. Consider the
following set of equations arisen from a modeling exercise:
(27)

(28)

With . Task is to find a solution for x(t) and y(t).

32 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

Taking Laplace Transform of eqns. (27) & (28) we obtain,

(29)

(30)

After rearrangement of the above we obtain,

(31)

(32)

Taking the inverse Laplace Transform we obtain,

(33)

(34)
y(t)

Definition of Transfer Function

Transfer function is defined as the ratio of the Laplace Transform of the output variable to the Laplace
Transform of the input variable assuming all the initial condition as zero.

Properties of Transfer Function

• The transfer function of a system is the mathematical model expressing the differential equation
that relates the output to input of the system.
• The transfer function is the property of a system independent of magnitude and the nature of
the
input.
• The transfer function includes the transfer functions of the individual elements. But at the same
time, it does not provide any information regarding physical structure of the system.
• The transfer functions of many physically different systems shall be identical.
• If the transfer function of the system is known, the output response can be studied for various
typesof inputs to understand the nature of the system.
• If the transfer function is unknown, it may be found out experimentally by applying known
inputs tothe device and studying the output of the system.

How you can obtain the transfer function (T. F.)

• Write the differential equation of the system.
• Take the L. T. of the differential equation, assuming all initial condition to be zero.
• Take the ratio of the output to the input. This ratio is the T. F.

Mathematical Model of control systems

A control system is a collection of physical object connected together to serve an objective. The
mathematical model of a control system constitutes a set of differential equation.

First Order Process

A first order process is a process whose output y(t) is modeled by a first order differential equation.
(35)

33 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

where, are input and output of the process espectively. If , then define the

following: and
Hence, the first order differential equation takes the following form:
(36)

At steady state condition , the equation can be re-written as

(37)

Subtracting eq. (36) from eq. (37), we obtain

(38)

Alternatively,
(39)

Where, and are the deviation forms of the input and output variables
of the process around the steady state, whose initial conditions are assumed to be the
following: .
Taking Laplace Transform of the eq. (III.39) we obtain,
(40)
Rearranging the above we obtain,
(41)

Gp(s) is called the transfer function of the process. Kp and τp are called as gain and time constant of the
process. The unit of gain is the ratio of the units of output to that of input, whereas the unit of time
constant is same as that of time.

Examples of a first order process

Mercury in glass thermometer
Assumptions:
Lumped - All thermal resistance resides in fluid film surrounding bulb
Parameter - All thermal capacity is in the mercury and it is uniform throughout
Model - The glass does not expand
- Thermometer is initially at steady state

34 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

H = film coefficient of heat transfer [=] BTU/ hr • ft 2•o F

A = Area of bulb [=] ft2
x = Fluid Temp [=] °F
y = Hg Temp [=]°F
C = Heat capacity of Hg [=] BTU/lb m oF
m = Mass of mercury [=] lbm
t = time [=] hr
It is not convenient to work in terms of absolute temperature, instead solve the problem in terms of
deviations
For t ≤ 0 Thermometer is at steady state

Taking the LaPlace Transform

Where
= Time Constant [=] hours
Step Responseof a first order system
To examine the response of the thermometer to a step input in temperature, we would note that

where: A = the magnitude of the step change in temperature [=] °F

Then:

we can rearrange this equation into the format

35 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

The solution requires that partial fractions are used, so:

Where:
C1 = A
C2 = -A
The solution is

Dynamic response of a first order process upon step change in input

Ramp response of a first order system

We know, the transfer function of a first order system is
Y(s)/X(s) = 1/(𝜏 s + 1)
Where ‗𝜏‗ is the time constant.
We know, the ramp input is X(t) = Aramp t
Where ‗Aramp‗ is the magnitude of ramp input.
Or
X(s) = Aramp/s2
Applying the above into the first order transfer function, we get

By partial fraction technique, we get

Taking Laplace inverse on both sides, we get

36 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

This expression is the response of first order system to a ramp input.

Note: Aramp= 1, for unit ramp input.

Pulse response of a first order system:

We know, the transfer function for a first order system is
Y(s)/X(s) = 1/(𝜏 s + 1)
We know, the pulse input is

Apulse [(1/s) – (e /s)]

Where tois the pulse time.
Substituting the above into the transfer function, we get

(1)
By translation of functions, we have

Now, taking Laplace inverse on equation (1)

At t < to, equation (1) becomes

Taking Laplace inverse, we get

At t > to, the solution to equation (1) becomes

Which on simplification yields

Effect of parameters on the response of First Order Process

Suppose two first order processes have same static gain but different time constants.

It indicates that Process 1 is faster than process 2 as the time constant of Process 1 is smaller than that
of Process 2. The responses of the processes for same unit step change in input are given in the figure
below:

Dynamic profile of two first order processes with same gain but different time constants

37 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

Since the gain of the processes are same, the ultimate response reaches the same value. On the other
hand, suppose two first order processes have different static gains but same time constants.

It indicates that Process 2 has higher static gain than Process 1. The responses of the processes are given
in the figure below:

Dynamic profile of two first order processes with different gain but same time constants.

We observe that the processes have same initial slope of response. Process 2 settles at a higher steady
state value due to its higher static gain.

Problem 1
A thermometer having a time constant of 0.5min is placed in a temperature bath and after thermometer
comes to equilibrium with the bath, the temperature of the bath is increased linearly at the rate of
1oC/min.
(i) What is the difference between the indicated and bath temperature at 0.25min and 3min
after the change in temperature begins?
(ii) What is the maximum deviation between the indicated and bath temperature and when does
it occur?
(iii) How many minutes does the response lag the input after long enough time is elapsed?

From the given statement, it is clear that a (unit) ramp input of magnitude 1oC/min [X(t) = t] to a
First order (thermometer) system.
We know, the unit ramp response of a first order system is

(i) At t = 0.25min, X(t) = 0.25 and Y(t) = 0.0533.

Therefore, X(t) – Y(t) = 0.1967 oC
At t = 3min, X(t) = 3 and Y(t) = 2.5012.
Therefore, X(t) – Y(t) = 0.4988 0C
ii) Maximum deviation between the indicated and bath temperature
= ARamp
= 1 * 0.5 = 0.5 0C

38 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

Dynamics of CSTR
Considering that a first order homogeneous reaction taking place in a CSTR, at unsteady state
operation, we have
Rate of reactant A entering = Rate of reactant A leaving + Rate of disappearance of A within the
reactor + Rate of accumulation of A within the reactor

νo Co = νoC1 + V (k C1) + V (dC1/dt)

Where Co and C1 are the entering and leaving concentrations of reactant A; Vis the volume of reactor;
νo is the volumetric flow rate of A into the reactor.
Or
Co = C1 + (k ) C1 +  (dC1/dt)
Where  = V/νo, is the space time or the time constant of the reactor.

Taking Laplace transform on both sides, we get

Co(s) = C1(s) + (k ) C1(s) +  [s C1(s)] = [(1 + k ) +  s] C1(s)
C1(s)/Co(s) = 1/ [(1 + k ) +  s]
C1(s)/Co(s) = R/(𝜏’ s + 1)
Where 𝜏’ = ( + k ) and R = 1/(1 + k ) , which are the time constant and theresistance for the
reactorprocess.
The above expression represents the transfer function a CSTR, at unsteady state.

Dynamics of Liquid level Tank

Consider tank of uniform cross sectional area A to which is attached a flow resistance R such as a valve,
a pipe, or a weir.

Assume that qO, the volumetric flow rate (volume/time) through the resistance, is related to the head h
by the linear relationship

(1)
A resistance that has this linear relationship between flow and head is referred to as a linear resistance.‖
A time-varying volumetric flow 4 of liquid of constant density p enters the [Link] system can be
analyzed by writing a transient mass balance around the tank:
Mass flow in - mass flow out = rate of accumulation of mass in the tank

In terms of the variables used in this analysis, the mass balance becomes

39 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(2)
Combining Eqs. (1) and (2) to eliminate qO(t) gives the following linear differential equation:

(3)
Initially, the process is operating at steady state, which means that dh/dt = 0 and we can write Eq. (3)
as

(4)
where the subscript s has been used to indicate the steady-state value of the variable.
Subtracting Eq. (4) from Eq. (3) gives

(5)
If we define the deviation variables as

(6)
Taking the transform of Eq. (6) gives

(7)
Equation (7) can be rearranged into the standard form of the first order lag to give

(8)
where τ = AR
In comparing the transfer function of the tank given by Eq. (8) with the transfer function for the
thermometer given by Eq. (7), we see that Eq. (8) contains the factor R.
The term R is simply the conversion factor that relates h(t) to q(t) when the system is at steady state.
For this reason, a factor K in the transfer function K/(τs + 1) is often called the steady-state gain.
If the transfer function relating the inlet flow q(t) to the outlet flow is desired, then from Eq. (1)

(9)
Subtracting Eq. (9) from Eq. (1) and using the deviation variable
Q, = q. - qo, gives

(10)
Taking the transform of Eq. (10) gives

(11)

40 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

Combining Eqs. (11) and (8) to eliminate H(s) gives

(12)
The steady-state gain for this transfer function is dimensionless, which is to be expected because the
input variable q(t) and the output variable qO(t) have the same units (volume/time).

Liquid level process with constant flow outlet

Another example of a transfer function that often arises in control systems may be developed by
considering the liquid level system shown in Fig.
The resistance is replaced by a constant-flow pump.
The same assumptions of constant cross-sectional area and constant density that were used before,
also apply here.
For this system, q(t) is now a constant; thus

(1)
At steady state, Eq. (1) becomes
qs - q0 = 0 (2)

Subtracting Eq. (2) from Eq. (1) and introducing the deviation variables
Q = q - qs
and
H = h - hs, gives

(3)

Taking the Laplace transform of each side of Eq. (3) and solving for H/Q gives

(4)
The transfer function, 1/As, in Eq. (4) is equivalent to integration. Therefore, the solution of Eq. (4) is

If a step change Q(r) = u(t) were applied to the system shown in Fig. the result is

(6)

The step response given by Eq. (6) is a ramp function that grows without limit. Such a system that
grows without limit for a sustained change in input is said to have non-regulation.

Systems that have a limited change in output for a sustained change in input are said to have regulation.

Mixing Process:
Consider the mixing process shown in Fig. in which a stream of solution containing dissolved salt flows
at a constant volumetric flow rate q into a tank of constant holdup volume V.
41 Department of Chemical Engineering
UCH2504 Process Instrumentation, Dynamics And Control

The concentration of the salt in the entering stream, x (mass of salt/volume), varies with time. It is
desired to determine the transfer function relating the outlet concentration y to the inlet concentration
X.
Assuming the density of the solution to be constant, the flow rate in must equal the flow rate out, since
the
holdup volume is fixed.
A transient mass balance for the salt; thus
Flow rate of salt in - flow rate of salt out = rate of accumulation of salt in the tank
Expressing this mass balance in terms of symbols give

(1)
• At steady state, Eq. (1) may be written
qxs - qys = 0 (2)
Subtracting Eq. (2) from Eq. (1) and introducing the deviation variables
X = x - xs ,
Y=y-ys
Give

Taking the Laplace transform of this expression and rearranging the result give

(3)
where τ = V/q.

LINEARIZATION
All the examples of physical systems discussed so far , including the liquid-level system have been
linear. Actually, most physical systems of practical importance are nonlinear.
Characterization of a dynamic system by a transfer function can be done only for linear systems For
dynamic analysis, non linear systems can be approximated to linear ones. A very important technique
for such approximation is illustrated by the following discussion of the liquid-level system of Fig.

42 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

We now assume that the resistance follows the square-root relationship

q0 = Ch1/2 (1)
where C is a constant.
For a liquid of constant density and a tank of uniform crosssectional area A, a material balance around
the tank gives

(2)
Combining Eqs. (1) and (2) gives the nonlinear differential equation

(3)
Owing to the presence of the nonlinear term h1/2, We cannot proceed, for which there is no simple
transform. This difficulty can be circumvented as follows.

By means of a Taylor-series expansion, the function q0(h) may be expanded around the steady-state
value hs; thus

where q0‘(hs) is the first derivative of q0 evaluated at hs, q0‖(hs) the second derivative, etc. Keeping only
the linear term, gives

(4)
Taking the derivative of q0 with respect to h in Eq. (1) and evaluating the derivative at h = hs, gives
q0‘(hs) = (1/2)Chs-1/2
Introducing this into Eq. (4) gives

(5)
where qos = q0(hs)
(R1)-1 =1/[Link]-1/2
Substituting Eq. (5) into (2) gives

(6)
At steady state the flow entering the tank equals the flow leaving the tank; thus
q0 = q0s, (7)
Introducing this last equation into Eq. (6) gives

(8)
Introducing deviation variables Q = q - qs and H = h - hs into Eq. (8) and transforming give

Where

However, in this case, the resistance R1 depends on the steady-state conditions around

43 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

which the process operates.

Graphically, the resistance R1 is the reciprocal of the slope of the tangent line passing through the
point (q0s, hs) as shown in Fig.

Furthermore, the linear approximation given by Eq. (4) is the equation of the tangent line itself. From
the graphical representation, it should be clear that the linear approximation improves as the deviation
in h becomes smaller.

First Order systems in series:

Non-Interacting System:
The outlet flow from tank 1 discharges directly into the atmosphere before spilling into tank 2 and the
flow through R1 depends only on h1 The variation in h2 in tank 2 does not affect the transient response
occurring in tank 1.
Assume the liquid to be of constant density, the tanks to have uniform cross-sectional area, and the flow
resistances to be linear.
The approach will be to obtain a transfer function for each tank,Q1(s)/Q(s) and H2(s)/Q1(s), by writing
a transient mass balance around each tank; these transfer functions will then be combined to eliminate
the intermediate flow Q1(s) and produce the desired transfer function.

A balance on tank 1 gives

(1)
A balance on tank 2 gives

(2)
The flow-head relationships for the two linear resistances are given by the expressions

(3)

44 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(4)
Steady state balance around tank 1 gives
qs – q1s = 0 (5)
q1s – q2s = 0 (6)
Subtracting (5) from (1) and using (3) gives

(7)
Subtracting (6) from (2) and using (4) gives

(8)
Introducing deviation variables, Q1 = q1 – q1s, Q = q - qs, and τ1= R1A1

(9)

(10)
Taking Laplace transform of Eqns (9) and (10) gives

(11)

(12)
where H2 = h2 - h2s, and τ2 = R2A2
The overall transfer function H2(s)/Q(s) can be obtained by multiplying Eqs. (11) and (12) to eliminate
Q 1(S):

(13)
Generalization for several Non-interacting systems in series:
The overall transfer function of Eq. (13) is the product of two first-order transfer functions, each one of
which is the transfer function of a single tank operating independently of the other.
We may now generalize by considering n non-interacting first-order systems as represented by the block
diagram of Fig.

Non-Interacting first order systems

The block diagram is equivalent to the relationships

45 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

To obtain the overall transfer function, we simply multiply together the individual transfer functions;
thus

Interacting system:
The flow through R1 depends on the difference between h1 and h2.

The analysis is started by writing mass balances on the tanks as was done for the non-interacting case.
The balances on tanks 1 and 2 are the same as before and are given by Eqs. (1) and (2). However, the
flow-head relationship for tank 1 is now

(14)
The flow-head relationship for R2 is the same as before and is expressed by Eq. (4).
At steady state, Eqs. (1) and (2) can be written
qs – q1s = 0 (15)
q1s – q2s = 0 (16)
Subtracting Eq. (15) from Eq. (1) and Eq. (16) from Eq. (2) and introducing deviation variables give

(17)

(18)
Where,
Q1 = q1 – q1s,
Q = q – qs,
Q2 = q2 – q2s,
H1 = h1 – h1s,
H2 = h2 - h2s,
Expressing Eqs. (14) and (4) in terms of deviation variables gives

(19)

46 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(20)
Transforming Eqs. (17) through (20) gives
Q(s) – Q1(s) = A1sH1(s) (21)
Q1(s) – Q2(s) = A2sH2(s) (22)
R1Q1(s) = H1(s) - H2(s) (23)
R2Q2(s) = H2(s) (24)
Substituting Eq (24) in Eq (22) gives

(25)
Substituting Eqs (23) and (25) in Eq (21) gives

The above equation can be simplified to produce the desired transfer function for the interacting
system.

(26)
The term interacting is often referred to as loading. The second tank of [Link] said to load the first
tank.

Second Order
Transfer Function for Damped vibrator
A block of mass ‗W‗ resting on a horizontal, frictionless table is attached to a linear spring. A viscous
damper (dashpot) is also attached to the block.
Assume that the system is free to oscillate horizontally under the influence of a forcing function F(t).
The origin of the coordinate system is taken as the right edge of the block when the spring is in the
relaxed or un-stretched condition. At time zero, the block is assumed to be at rest at this origin.
the right of Y = 0 and when it is moving toward the
right (positive direction).
Under these conditions, the position Y and the velocity dF/dt are both positive.
At this particular instant, the following forces are acting on the block:
The force exerted by the spring (toward the left) of -KY where K is a positive constant, called Hooke‗s
constant.
The viscous friction force (acting to the left) of -C dY/dt, where C is a positive constant called the
damping coefficient.
The external force F(t) (acting toward the right).

(1)

(2)
Where W= mass of block lbm,
gc = 32.2 (lbm)(ft)/(lbf)(sec2)

47 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

C = viscous damping coefficient , lbf/ (ft/sec)

K= Hookes constant , lbm/ft
F(t) = driving force a function of time lbf

(3)

(4)

(5)

(6)

(7)

(8)

(9)
If the block is motionless (dY/dt = 0) and located at its rest position (Y= 0) before the forcing function
is applied , The laplce transform of eq.4 becomes

The transfer function

(11)

Dynamic Response of a Second Order Process to a Step Change in the Input

For a step input of magnitude A , the Laplace Transform of u(t) would be,
(12)

Hence, second order process takes the following form,

(13)

The process response will grossly depend upon the value of ξ and there can be three distinguished cases
of ξ, i.e. ξ >1; ξ = 1 and ξ <1 .

Case A: ξ = 1
In this case the process response equation in the Laplace domain takes the following form:

48 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(14)

Using the following:

(15)

in eq. (14), we obtain

(16)

or

(17)
For ξ ≠ 1, using the following:

(18)

in eq. (13), we obtain

(19)

Case B: When ξ >1

(19
)

49 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

In the above equations, the following trigonometric identities have been used:

and .
Hence we get the final expression for process response when ξ >1,
(20)

Case C:

(21
)

In the above equations, the following trigonometric identities have been used:
(22)

(23)

One can also use the following trigonometric identity for the above expression:
(24)

Hence,
(25
)

Hence we get the final expression for process response for ξ < 1,
(26)

The frequency of oscillation is

(27)

whereas the phase lag is

50 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(28)

Features of the process response

Let us find out the initial and final values as well as initial and final slopes of second order processes.
For
(29)

Hence,
(30)

For
(31)

Hence,

(32)

For either value(s) of ,

( 33)

he figure below is the graphical representations of three cases of discussed so far.

Step response of a second order process

For and , the process response never goes beyond the final steady state value, whereas
for , the response exceeds the ultimate steady state several times during its transient motion. In
other words, some oscillatory behavior is observed in this case. Nevertheless a damping action is active
in all the three cases which is high for and low for . Hence, is called the damping

51 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

coefficient . The response is overdamped if , critically damped if and underdamped if

.
Characteristics of an underdamped process
Following figure is a graphical representation of a typical second order underdamped process.

Step response of a second order underdamped process

Suppose P is the maximum amount by which the underdamped response overshoots its ultimate steady
state value and Q is the ultimate steady state value. Then the ratio is called overshoot . Suppose R is
the second largest value by which the response exceeds its ultimate steady state value. Then the
ratio is called the decay ratio . In general, decay ratio is a measure of how rapidly the oscillations
decrease. It is to be noted that the ratio of two successive peaks is a constant quantity and is equal to
the decay ratio for that underdamped process.
Following are the important features of an underdamped process:
Period of Oscillation : Time elapsed between two successive peaks is called the period of oscillation
(T ). Hence,
(101)

Overshoot : The slope at every zenith and nadir of the oscillation would be zero. The first zenith
indicates the value of . The first zenith appears after a time of . Hence,

(102)

and
(103)
Hence,

52 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

(104)

and,
(105)

Decay Ratio : The second zenith indicates the values of and that appears at a time .

(106)

Hence,
(107)

and,
(108)

Hence,
(109)

Problem 1
Derive the transfer function H/Q for the liquid level system shown below; [H and Q are the deviation
variables in h (m) and q (m3/min) respectively, Area A = 0.75m2].

Writing the material balance for the given liquid level system, we get
q = h [(1/R1) + (1/R2)] + A (dh/dt) ------------ (1)
At steady state, equation (1) becomes
53 Department of Chemical Engineering
UCH2504 Process Instrumentation, Dynamics And Control

qs = hs [(1/R1) + (1/R2)] + A (dhs/dt) ------------ (2)

Note: dhs/dt = 0
Subtaracting (2) from (1) gives
q - qs = [h - hs] [(1/R1) + (1/R2)] + A [d(h - hs)/dt] Or Q = H [(1/R1) + (1/R2)] + A [dH/dt]
Taking Laplace transform on both sides, we get
Q(s) = H(s) [(R1+R2)/(R1R2)] + A [s H(s) – H(0)] = H(s) [(R1+R2)/(R1R2)] + A [s H(s)];
since H(0) = hs – hs = 0
Or
H(s)/Q(s) = Rtotal/( total s + 1) ------------- (3)
Where Rtotal = [R1R2)/(R1+R2)] and
total = A Rtotal Equation (3) represents the transfer function for the given liquid level system.
Now, substituting the given values into equation (3) gives
H(s)/Q(s) = 0.6667/(0.5s + 1)

Problem 2

A liquid level process is operating at steady state and the input to the tank is 10m3/min. At t = 0, 1m3 of
water is added suddenly (unit impulse) to the tank; at t = 1, 2m3 of water is added suddenly to the tank.
The resistance (R) at the exit is 0.5 and the time constant,  = 1. Plot the response of the level in the
tank and determine the level at t = 0.5, 1 and 1.5
Ans.
Given, Liquid level process (first order) – subjected to two different impulse input at the time interval
of 1 min.
Steady flow rate, qs = qo s = 10 m3/min
Resistance, R = 0.5 min/m2 and
Time constant,  = 1 min
We know, the transfer function for a first order liquid level system is
H(s)/Q(s) = R/( s + 1)
Sub. the given data‗s into the above, we get
H(s)/Q(s) = 0.5/(s + 1) ----------- (1)
➢ At t = 0, Q(t) = 1 (impulse) or Q(s) = 1 ----------- (2)

Sub. (2) into (1), we get

H(s) = 0.5/(s + 1)
Or
H(t) = 0.5 e-t ------------------(3)
Or
h(t) = hs + 0.5 e-t, 0 < t < 1 ---------- (4)
We know, at steady state,
qo s = hs/R or hs = qo s * R
Sub. the given data‗s into the above, we get
hs = 10 * 0.5 = 5 m Equation (4) becomes
h(t) = 5 + 0.5 e-t, 0 < t < 1 ----------- (5)
➢ At t = 1, Q(t)t → t - 1 = 2 (impulse) or Q(s) = 2----------- (6)

Sub. (6) into (1), we get

H(s) = 1/(s + 1)
Or
H(t)t → t - 1 = e-(t - 1)
Or
H(t) = 2.7183 e-t, ----------- (7)
54 Department of Chemical Engineering
UCH2504 Process Instrumentation, Dynamics And Control

The overall response is given by adding equations (3) and (7).

That is, H(t) = 0.5 e-t + 2.7183 e-t = 3.2183 e-t
Or
h(t) = 5 + 3.2183 e-t, t > 1------------ (8)
Tabulation: [By using equations (5) and (8)]
Now,

Now, plot the graph between h(t) and ‗t‘. Note: the bold numbers in the tabular column indicates the
answer for the second part.

Problem 3
A process of unknown transfer function is subjected to a unit-impulse input. The output of the process
is measured accurately and is found to be represented by the function y(t) = t e-t. Determine the unit
step response of the process.
Given, Unknown transfer function – subjected to unit-impulse input.
The output of the process is Y(t) = t e-t
We know, the unit-impulse input is
X(t) = 1 or
X(s) = 1
Taking Laplace inverse on the given output, we get
Y(s) = 1/(s + 1)2
From the above, we get the transfer function as
Y(s)/X(s) = 1/(s + 1)2
Now, we know the unit step input is
X(t) = 1 or X(s) = 1/s
Substituting this X(s) into the above transfer function, we get
Y(s) = 1/[s (s + 1)2]
By partial fraction technique, the above equation becomes
Y(s) = (1/s) – [1/(s + 1)] – [1/(s + 1)2]
Taking Laplace inverse on both sides, we get Y(t) = 1 – e-t – t e-t = 1 – (1 + t) e-t

Problem 4
A step change of magnitude 4 is introduced into the system having the transfer function;

Determine (i) Percent overshoot (ii) Decay ratio (iii) Period of oscillation (iv) Rise time and (v) Settling
time.
Ans:
G(s) = Y(s)/X(s) = 10/(s2 + 1.6s + 4)
Writing the given second order transfer function in its standard form as
Y(s)/X(s) = 2.5/[0.25s2 + 0.4s + 1] --------- (1)
On comparing equation (1) with the standard second order transfer function,
we get 2 = 0.25 or = 0.5
And 2 ξ  = 0.4 or ξ = 0.4/2 = 0.4/(2*0.5) = 0.4 or ξ < 1
Therefore, the given second order system is under damped.
(i) Overshoot = exp [(-π ξ)/√(1 – ξ2)]

55 Department of Chemical Engineering

UCH2504 Process Instrumentation, Dynamics And Control

Sub. ξ = 0.4 into the above, we get Overshoot = 0.2538 or 25.38%

(ii) Decay ratio = exp [(-2 π ξ)/√(1 – ξ2)] = (Overshoot)2

Sub. overshoot = 0.2538 into the above, we get Decay ratio = 0.0644
(iii)Period of oscillation, T = 2π/ω = (2 π )/√(1 – ξ2)
Sub.  = 0.5 and ξ = 0.4 into the above, we get T = 3.4278 time/cycle
(iv) Rise time, tr = (π – φ)/ω
Where, ω = √(1 – ξ2)/  and φ = tan-1[√(1 – ξ2)/ξ] Sub.  = 0.5 and ξ = 0.4 into the above, we get ω =
1.833 rad/min and φ = 1.1593 rad
Then, Rise time, tr = 1.0815 Now,
Given (step input) X(t) = 4 or
X(s) = 4/s----------- (2) Sub. equation (2)
into (1), we get
Y(s) = 10/[s (0.25s2 + 0.4s + 1)]
Where Ultimate value, Yult = 10
We know, the step response of an under damped second order system is
Y(t) = Yult {1 – [e- ξt//√(1 – ξ2)] sin(ωt + φ)}
Sub. the known values into the above, we get
Y(t) = 10 {1 – [(1.0911 e-0.8t) sin(1.833t+1.1593)]} ----------- (3)
(v) Response time (tR): It is the time requires for the response to come within ± 5 percent of its
ultimate value and remain there.
At t = tR, Y(t) = 0.95 Yult = 0.95 * 10 = 9.5 ----------- (4)
Sub. equation (4) into equation (3), we get e-0.8tR sin(1.833t+1.1593) = 0.04583
This on solving by trial and error method gives tR ≈ 3.8

56 Department of Chemical Engineering