Dynamic Behavior

In analyzing process dynamic and process control systems, it is

important to know how the process responds to changes in the
process inputs.
A number of standard types of input changes are widely used for
two reasons:

1. They are representative of the types of changes that occur

in plants.

2. They are easy to analyze mathematically.

1
1. Step Input

A sudden change in a process variable can be approximated by

a step change of magnitude, M:

 0 t0 M
us (t )   Us s 
M t  0 s

The step change occurs at an arbitrary time denoted as t = 0.

• Special Case: If M = 1, we have a “unit step change”. We

give it the symbol, u(t).

• Example of a step change: A reactor feedstock is suddenly

switched from one supply to another, causing sudden
changes in feed concentration, flow, etc.
2
Example:
The heat input to the stirred-tank heating 10,000 kcal/hr

system is suddenly changed from 8000 to

8,000 kcal/hr
10,000 kcal/hr by changing the electrical
signal to the heater. Thus,
and q t   8000  2000u t  , u t  : unit step
q ' t   q  q  2000u t  , q  8000 kcal/hr
In the Laplace domain, the step input has the form:
2000
Q'(s)=
s

3
2. Ramp Input
• Industrial processes often
experience “drifting disturbances”,
that is, relatively slow changes up
or down for some period of time.
• The rate of change is
approximately constant.
We can approximate a drifting disturbance by a ramp input:
0 t  0 a
uR  t    UR s 
at t  0 s2
Ramp a setpoint to a new value. (Why not make a step change?)

Examples of ramp changes: Feed composition, heat exchanger

fouling, catalyst activity, ambient temperature.
4
3. Rectangular Pulse

It represents a brief, sudden change in a process variable:

XRP

0 for t  0 h

u RP  t   h for 0  t  t w
0 for t  tw Tw Time, t 0


h
U RP  s   (1  e-tw s )
s
Examples:

1. Reactor feed is shut off for one hour.

2. The fuel gas supply to a furnace is briefly interrupted.
5
Common process inputs:

M
Us  s 
s

a
UR s 
s2

h
U RP  s   (1  e-tw s )
s

6
4. Sinusoidal Input

Processes are also subject to periodic, or cyclic, disturbances.

They can be approximated by a sinusoidal disturbance:

0 for t  0 A
usin  t    U sin  s  
 A sin t  for t  0 s2   2

where: A = amplitude,  = angular frequency

Examples:

1. 24 hour variations in cooling water temperature.

2. 60-Hz electrical noise (in the USA)

7
5. Impulse Input

• Here, u I  t     t 
• It represents a short, transient disturbance.
Examples:

1. Electrical noise spike in a thermo-couple reading.

2. Injection of a tracer dye.
• Useful for analysis since the response to an impulse input
is the inverse of the TF. Thus,
 t  y t 
G s
1 Y s
Here, Y  s   G  s   s 

y (t )  L1 Y  s    L1 G  s    s    g (t )

8
First-Order Systems qi

The standard form for a first-order TF is:

Y s  K
 (1)
U  s  τs  1
where:
K  steady-state gain q
τ  time constant
Consider the response of this system to a step of magnitude, M:

M
u  t   M for t  0  U s 
s
Substitute into (1) and rearrange,
KM
Y s   (2)
s  τs  1
9
Take L-1,


y t   KM 1  e t /τ 
Let y = steady-state value of y(t).
From the previous equ. y  KM .

y

Large τ means
a slow
response

10
 Representing First-Order Systems in MATLAB

>> K=1;
>> M=2;
>> tau=0.1;
>> g=tf(K*M, [tau 1])

Transfer function:
2
---------
0.1 s + 1

>> step(g)

11
Integrating Process
Not all processes have a steady-state gain. For example, an
“integrating process” or “integrator” has the transfer function:

Y s  K
 K  constant 
U s  s

Consider a step change of magnitude M. Then U(s) = M /s and,

KM L-1
Y s   2
 y t   KMt
s
Thus, y(t) is unbounded and a new steady-state value does not
exist.
g=tf(K*M, [ 1 0 0])
12
Common Physical Example:
Consider a liquid storage tank with a pump on the exit line:

- Assume:
1. Constant cross-sectional area, A.
2. q  f  h 
dh
- Mass balance: A  qi  q (1)  0  qi  q (2)
dt
- Eq. (1) – Eq. (2), take L, assume steady state initially,
H  s  
1
Qi  s   Q  s   H  s 1
As 
Qi  s  As
1
- For Q    (constant q),
s  0 H   
s  Qi  s 
As 13
Second-Order Systems
• Standard form:
Y s  K

U s  τ 2s 2  2ζτs  1

which has three model parameters:

K  steady-state gain (process gain)

τ  "time constant" [=] time
ζ  damping coefficient (dimensionless)

• Equivalent form: divide num. and den. by (τ2):

Y s  K n2  1
  where   natural frequency  
U s 
n
s 2  2ζn s  n2  τ

14
• The type of behavior that occurs depends on the numerical
value of damping coefficient, ζ :
It is convenient to consider three types of behavior:
Damping Type of Response Roots of Characteristic
Coefficient Polynomial
ζ 1 Overdamped Real and ≠
ζ 1 Critically damped Real and =
Underdamped Complex conjugates (with
0  ζ 1
negative real parts)
ζ0 Critically stable Complex (0 real parts)
Complex conjugates (with
ζ0 Unstable positive real parts)

• Note: The characteristic polynomial is the denominator of the

transfer function: τ 2 s 2  2ζτs  1
15
 Step Response of Second-Order Systems

The Laplace domain response of the system Y(s) for a step

input can be written as: M K
Y (s ) 
s  2s 2  2s  1

The quadratic term may be factored into two terms that

contain the roots
 2
 1
sa   
 
 2 1
sb   
 

Then we can rewrite Y(s) as:

KM /  2
Y (s ) 
s s  sa s  sb 

16
The response of the system y(t) can be
found by inverting this equation. The
roots sa and sb will be real or complex
depending on value of the parameter  .
The nature of the roots will, in turn,
affect the form of y(t).

CASE (I) STEP RESPONSE WHEN 0<  < 1.


      
   1 - 
2
   1 -  2  

y(t )  KM  t  
-t /
1 - e  cos  t  sin 
        


   1- 2
  

 
CASE (II) STEP RESPONSE WHEN  = 1.
  t  t / 
 
y(t )  KM 1  1   e 
    
 
CASE (III) STEP RESPONSE WHEN  > 1.
      
   1  2    1  2  
y(t )  KM 1  e t  
t /  
 cosh  t  sinh 
        

  1 2
  
 
 17
 Step response of second-order processes
MATLAB simulation
clc; close all;
K=10;
M=2;
tau=[0.1];
underdampd z=1;
g=tf(K*M, [ tau^2 2*z*tau 1])
step(g/K*M)
title('Step to a second order
critically-damped TF');
xlabel('Time')
ylabel('Process Output / KM')

overdampd
SIMULINK simulation

18
 General Remarks

Several general remarks can be made concerning the

responses of 2nd order system show :

1. Responses exhibiting oscillation and overshoot (y/KM > 1) are

obtained only for values of (ζ<1)
2. Large values of ζ yield a sluggish (slow) response. (ζ>1)
3. The fastest response without overshoot is obtained for the
critically damped case  ζ  1 .

19
 Performance characteristics for the step response of
an underdamped process

Overshoot: OS = a/b P

Decay Ratio: DR = c/a a c 1.05b

y(t )
Rise Time: tr KM 0.95b

Time to First Peak: tp b

Settling Time: ts
Period of Oscillation: P tr tp ts

20
 Terms Used to Describe an Underdamped System

1. Rise Time: tr is the time the process output takes to first reach the new
steady-state value. Some textbooks have another definition: the time the
process output takes to rise from 10% to 90% of the step size.
2. Time to First Peak: tp is the time required for the output to reach its first
maximum value.
3. Settling Time: ts is defined as the time required for the process output to
reach and remain inside a band whose width is equal to ±5% of the total
change in y. The term 95% response time sometimes is used to refer to this
case. Also, values of ±1 and ±2% sometimes are used.
4. Overshoot: OS = a/b (% overshoot is 100a/b).
5. Decay Ratio: DR = c/a (where c is the height of the second peak).
6. Period of Oscillation: P is the time between two successive peaks or two
successive valleys of the response.

21
Analytical expressions can be developed for some of the TF
Characteristics, examples are:

Time to First Peak (tp): t p   / 1   2

Overshoot (OS):

Decay Ratio (DR):

Period of Oscillation P  2 / 1   2

22
 Example
COOLANT OUT Product

TT T

TC

REACTANT, FO
COOLANT IN,
F C , TC

A stirred-tank reactor is equipped with a cooling jacket to remove

heat liberated in the reaction. A proportional controller is used to
regulate coolant flow rate so as to keep the reactor temperature
reasonably constant. The controller has been designed so that the
controlled reactor exhibits typical underdamped second-order
temperature response characteristics when it is disturbed, either by
feed flow rate or by coolant temperature changes.
23
 Example
(a) The feed flow rate to the reactor changes suddenly from
0.4 to 0.5 kg/s, and the temperature of the reactor contents,
initially at 100 °C, changes eventually to 102 °C. What is
the gain of the transfer function (under feedback control)
that relates changes in reactor temperature to changes in
feed flow rate? (Be sure to specify the units.)
Temp (C)

102.5C
P

a 102.03C
y T
K   c New SS
102C
u Fo
102  100 
C
  20 b
0.5  0.4 kg / s
Initial SS
100C
0 1000 3060 time (s)

24
 Example
(b) The operator notes that the resulting response is slightly oscillatory with
maxima estimated to be 102.5 and 102.03 °C occurring at times 1000
and 3060 s after the change is initiated. What is the complete process
transfer function?
2
ln OS 
  
2
2  ln OS 
 
102.5  102
OS   0.25
102  100
  0.4037  0.4

1  2
P  2 / 1   2  P
2
P  3060  1000  2060 s
  300 s
T s  K 20
Hence the process TF is:  
Fo  s  2 2
τ s  2ζτs  1 90000s 2  240s  1
25
 Example

(c) The operator failed to note the rise time. Predict tr , based on the results in
(a) and (b). 
     

 -t /    1 -  2    1 -  2  
Using the time domain y(t )  KM 1 - e  cos  t  sin  t 
       


    1 -  2
 

solution for the  
underdamped case (0<<1),     
-t /    1 -  2    1 -  2 
0e  cos  t  sin  t 
setting y(tr)=KM, then the     1 -  2 
  
  
bracketed quantity will be
   
equal to zero at t=tr:  1   2    1   2 
0  cos  tr   sin  tr 
 
   1  2   
Solving for the rise time we get:

tr    cos  1

syms tr1 1  2
z=0.4; tau=300;
f=cos((sqrt(1-z^2)/tau)*tr1)+(z/sqrt(1-z^2))*sin((sqrt(1-z^2)/tau)*tr1);
tr=double(solve(f==0,tr1))

Substituting for  =0.4 and =300 s we get: tr = 649 s

26
 Representing Second-Order Systems in MATLAB

>> K=1;
>> M=2;
>> tau=0.1;
>> zeta=0.5;
>> g=tf(K*M, [tau^2 2*zeta*tau 1])

Transfer function:
2
--------------------
0.01 s^2 + 0.1 s + 1

>> step(g)

27
Effect of varying (
K=1; M=2; tau=0.01;
ζ)
zeta=0.1; g1=tf(K*M, [tau^2 2*tau*zeta 1]);
zeta=0.5; g2=tf(K*M, [tau^2 2*tau*zeta 1]);
zeta=1; g3=tf(K*M, [tau^2 2*tau*zeta 1]);
zeta=3; g4=tf(K*M, [tau^2 2*tau*zeta 1]);
step(g1,g2,g3,g4)
OR
K=1; M=2; tau=0.01;
for zeta=[0.1 0.5 1 3]
g=tf(K*M, [tau^2 2*tau*zeta 1]);
step(g); hold on
end

Effect of varying ( )
K=1; M=2; zeta=1;
tau=0.1; g1=tf(K*M, [tau^2 2*tau*zeta 1]);
tau=0.2; g2=tf(K*M, [tau^2 2*tau*zeta 1]);
tau=0.3; g3=tf(K*M, [tau^2 2*tau*zeta 1]);
tau=0.4; g4=tf(K*M, [tau^2 2*tau*zeta 1]);
step(g1,g2,g3,g4)
OR
K=1; M=2; zeta=1;
for tau=[0.1 0.2 0.3 0.4]
g=tf(K*M, [tau^2 2*tau*zeta 1]);
step(g); hold on
end
28
 Response of two interacting tanks

Y s  K

U s  τ 2s 2  2ζτs  1

  1 2  (A1R1 )(A 2 R 2 )

1   2  A1R 2
 
2
R1=10; R2=10;
A1=1; A2=1;
K=R2; M=2;
tau=sqrt(A1*A2*R1*R2);
zeta=(A1*R1+A2*R2+A1*R2)/(2*tau)
tau=0.1; zeta=1; g=tf(K*M, [tau^2 2*tau*zeta 1]);
step(g)
29