Dynamic Behavior of First-Order and Second-Order Processes: CAB3014 - Chemical Process Dynamics and Control

Process Dynamics and Control
Chapter 5
Dynamic Behavior of First-order

and Second-order Processes
07/25/20 CAB3014 - Chemical Process Dynamics and Control 1

Chapter Objectives
End of this chapter, you should be able to:
1. Determine dynamic behavior of chemical processes which

are classified as First-order and Second-order processes
2. Explain various test inputs used for dynamic analysis of
systems
3. Derive expressions for transient response for different
inputs

Introduction
• Investigate how processes respond to

typical changes in their environment.
• Transfer function representation makes it
easy to compare the effects of different
inputs.
• The dynamic behavior of a given process
can be generalized easily.

Introduction
• For a general first-order transfer function with output Y(s)

and input U(s),
K
Y (s)  U ( s)
s  1
a general time-domain solution can be found once the

nature of the input change is specified.
• The solution applies to any other process with a first-
order transfer function.
• We will consider several standard process inputs that
are used to characterize the behavior of many actual
processes.

Standard Process Inputs
2 types of inputs:
What are they?

• Inputs that can be manipulated to control

the process
And
• Inputs that are not manipulated, classified

as disturbance variables

• The word input refers

to any variable that
influences the process
output.
• Step input (Sudden
and sustained input
change)
0 t  0
U S (t )  
M t  0
M
U S ( s) 
s

Ramp input
• Ramp input (gradual upward or downward

change)
0 t  0 a
 U R (s)  2
U R (t )   s
at t  0

• Rectangular pulse (a step change limited to
certain period of time)
0 t  0
 h
U RP (t )  h 0  t  tw U RP ( s )  (1  e tws )
0 s
 t  tw

Sinusoidal input
• Sinusoidal input
0 t  0 A
U sin (t )   U sin ( s ) 
 A sin t t  0 s2   2

Impulse input
• Impulse input,  (t ) - Dirac delta function

L (t )  1
• Random inputs – changes in a complex manner –

Characterization will be in terms of statistical
terms such as mean and standard deviation.

Response of first-order processes
• We use the general first-order transfer

function
K
Y ( s)  U (s) (5.1)
s  1
where K is the process gain and τ is the time constant
Step response
For a step input of magnitude M, U ( s)  M / s , and
Eq.(5.1) becomes
KM
Y (s)  (5.2)
s s  1

Step Response of 1st order system
• The time-domain response is given by
y (t )  KM (1  e  t /  ) (5.3)
t y(t)
0 0
1 0.6321
2 0.8647
3 0.9502
4 0.9817
5 0.9933

Ramp response
a Ka
U R (s)  Y (s)  (5.4)
s2 s 2 (s  1)
Expanding into partial fractions, we get

Ka A B c
Y (s)     2 (5.5)
s (s  1) s  1 s s
2
The Heaviside expansion gives

Ka 2 Ka Ka (5.6)
Y (s)    2
s  1 s s

Taking the inverse Laplace transform of eq. (5.6) we get
y (t )  Ka (e  t /   1)  Kat (5.7)

Sinusoidal response

usin (t )  A sin t U ( s) 
s2   2
For a first-order process

Kp  0 1s 2
Y ( s)     
s  1 s 2   2 s  1 s 2   2 s 2   2
By partial fraction decomposition
KA 2  KA KA

0  2 2 1  2 
  1  2 2  1  2 2  1
Inverting, we get
KA t 
y (t )  (e   cos t  sin t )
  1
2 2

• Using the trigonometric identity
p cos A  q sin A  r sin( A   )
p
where r p2  q2 and tan  
q
We have This term dies out for large t

K p t 
Kp
y (t )  e  sin(t   )
  1
2 2
  1
2 2
where    tan 1 ( )

Note:  is not a function of t but of τ and .
For large t, y(t) is also sinusoidal, output sine is attenuated

by… 1
 2 2  1
Example: Electrically heated stirred-tank
• Consider a stirred-tank heating system with

constant holdup.
Ti
wi
T, w
V
Q
Suppose that the metal heating element has a significant

thermal capacitance and that the electrical heating rate
Q directly affects the temperature of the element rather
than the liquid contents.
Also assume that the element has a uniform
temperature, Te. (interpreted as the average
temperature for the heating element).
The unsteady-state energy balance for the tank and the

heating element can be written as
dT
mC  wC (Ti  T )  he Ae (Te  T ) (1)
dt
dTe
me C e  Q  he Ae (Te  T ) (2)
dt
where m  V and me C e is the product of the mass of the

metal in the heating element and its specific heat.
Q is the thermal equivalent of the instantaneous electrical

power dissipation in the heating element.
• The input variables are : w, Ti and Q
Under steady-state conditions, we have

0  wC (Ti  T )  he Ae (Te  T ) (3)
0  Q  he Ae (Te  T ) (4)
Subtracting (3) from (1) and (4) from (2), we have
 wC  (Ti  T ) i  (T  T )  he Ae  (Te  Te )  (T  T )
dT
mC
dt (5)
 (Q  Q )  he Ae  (Te  Te )  (T  T )
dTe
me C e
dt (6)
Substitute deviation variables, multiply (5) by 1/wC and
(6) by 1/heAe:
m dT  hA
 (T   Ti)  e e (Te  T ) (7)
w dt wC
me Ce dTe Q
  (Te  T ) (8)
he Ae dt he Ae
Laplace transform each equation, and after

rearrangement, is
m hA  hA
 s  1  e e T ( s )  Ti( s )  e e Te( s ) (9)
w wC  wC
 me Ce  Q( s ) (10)
 s  1Te( s )   T ( s )
 he Ae  he Ae

Ti
Eliminate Te(s) from (9) by using (10).

 me C e 
Multiply (9) by  s 1 and substitute to obtain
 he Ae 
 m m e C e 2  m e C e me C e m   m C 
 s      s  1T ( s)   e e s  1Ti( s)  1 Q ( s ) (11)
w h A hA wC w   hA  wC
 e e  e e   e e 
The transfer function relating T  to Q is

T ( s ) 1 / wC
  G1 ( s )  Ti( s)  0 (12)
Q ( s ) b2 s  b1 s  1
2
and the transfer function relating T  to Ti is

 me C e 
 s  1
T ( s )  he Ae
   G ( s)  Q ( s)  0 (13)
Ti( s) b2 s 2  b1 s  1
2

mC mC m m me C e
where b1  e e  e e  and b2 
he Ae wC w w he Ae
By the principle of superposition

T ( s )  G1 ( s )Q ( s )  G2 ( s)Ti( s ) (14)
Factoring the characteristic polynomial,
b2 s 2  b1s  1 as  1 s  1 2 s  1
We have two time constants.

A limiting behavior of has me Cand and
e 0 b2  0
b1  m / w(14) simplifies to
1 / wC 1
T ( s )  Q( s )  Ti( s ) (15)
m m
s 1 s 1
w w

Problem
The stirred-tank heating system is used to preheat a

reactant containing a suspended solid catalyst at a
constant flow rate of 1000kg/h. The volume in the
tank is 2 m3, and the density and the specific heat of
the suspended mixture are, respectively, 900 kg/m3
and 1 cal/g oC. The process initially is operating
with inlet and outlet temperatures of 100 and 130oC.
The following questions concerning the process
operation are posed:


1. What is the heater input at the initial steady state

and the values of K and  ?

2. If the heater input is increased by 30%, how
long will it take for the tank temperature to
achieve 99% of the final temperature change?
3. Assume the tank is initially at its steady state. If
the inlet temperature is increased suddenly from
100 to 120 oC, how long will it take before the
outlet temperature changes from 130 to 135 oC?

SOLUTION
1. From the steady-state energy balance, we have

Q  w C (T  Ti )  (106 )  (1)  (130  100)  3x107 cal/h.
O
1 1 6 C
K   10
wC (10 6 )  (1) cal/h
V (2)  (9x10 5 )
  6
 1.8h
w 10

2. 99% of the response is obtained in 5 process time
constants, i.e. 5 x 1.8 = 9 h.

The steady-state change in the temperature is
 10 6 9  10 6 
T (t  )  lim s   9 o C.
s 0 1.8s  1 s
 
The new steady-state outlet temp. is 130+9 = 139 oC.
3. The gain of the transfer function relating T  to Ti is

one, a change of 20oC will cause a change of 20oC in
the outlet.
y (t ) 5
 1  e t /   1  e  t / 1 .8 t = 0.52 h.
KM (1)(20)

So far in this chapter
• Standard test signals and their Laplace

transforms
• First order system and its response to
Step, ramp and sinusoidal inputs
• Example – Stirred tank heater system
• Numerical example

Today…
• Second order systems

• Classification of second order systems
• Step responses
• Characteristics of underdamped response

Response of second-order processes
A second-order transfer function can arise physically

whenever two first-order processes are connected in series.
K1 K 2 K
G(s) =  (16)
 1 s  1 2 s  1  1 2 s 2 + ( 1   2 )s + 1
 1 and  2, the first-order time constants, determine the

speed of response.

Second-order processes
A second-order transfer function will arise upon

transforming either a second-order differential equation
process model or two-coupled first-order equations.
K (17)
G(s) = 2 2
 s  2s  1
 (zeta) - damping coefficient – dimensionless –
measure of the amount of damping in the system.
Small  - little damping, large oscillations and vice versa.
 - natural period of oscillation = 1 /  n where  n is

the undamped natural frequency.

Classes of second-order systems
Damping Characterization Roots of

coefficient of response characteristic equation
> 1 Overdamped Real and unequal

= 1 Critically damped Real and equal
0 < < 1 Underdamped Complex conjugate
(  <0 is omitted because they are unstable systems).

The overdamped and critically damped forms of the second-
order transfer function most often appear when two first-
order systems occur in series as shown:
Transfer functions in (16) and (17) differ only in the form

of the denominators.
Equating the denominators yields

 2 s 2  2s  1   1s  1 2 s  1 (18)
1   2
where    1 2 and  =
2  1 2

Left side of (18) can be factored:
    
 s  2  1 
2 2  s 1 s  1
    2 1     2  1  (19)
  
from which we have

 
1  and 2  (20)
   2 1    12
After inverting to time domain, the responses are

categorized into three classes:

2nd-order process behavior under
Step response
For the step input (U(s) = M/s) to a process described by

(17)
KM
Y ( s) 
s  2 s 2  2s  1
(21)
After inverting to the time domain, the responses can be

categorized into three classes:

Overdamped    1
If the denominator of (21) can be factored using (20),

then the response can be written as
  1e t /1   2 e  t / 2 
y (t )  KM 1   (22)
 1   2 

Critically damped   1
  t  t /   (23)
y (t )  KM 1  1  e 
   

Underdamped 0    1
   1 2    1 2  
 
y (t )  KM 1  e t /  cos t  sin t  
    
 1 2  

 
 
 
(24)
• Using the trigonometric identity
p cos A  q sin A  r sin( A   )
p
where r p2  q2 and tan  
q

 1  1 2 

y (t )  KM 1  e t / 
sin  t   
 1  2   
  

Process Dynamics and Control Plots of responses for different values of 

Remarks:
Responses exhibit a higher degree of oscillation and

overshoot as approaches  zero.
Large values of  yield a sluggish (slow) response.
The fastest response without overshoot is obtained for


=1.

Characteristics of a step response
underdamped second-order process:

1. Rise time: tr, Time to first reach the new steady-state value.
2. Time to first peak: tp, Time to reach its first maximum value.
3. Settling time: ts, Time to reach and remain inside a band of

+5% of the total change.
4. Overshoot: OS=a/b
5. Decay ratio: DR = c/a
6. Period of oscillation: P is the time between two successive
peaks.
Analytical expressions for the
characteristics:
Time to first peak: t p  

(25)
1 2
  
Overshoot: OS  exp  
 1 2  (26)
 
 2 
Decay ratio: 
DR  exp   (OS) 2 (27)
 1 2 
 
2
Period: P (28)
1  2

2nd-order process behavior under
Sinusoidal response
The steady-state response to a sinusoidal input A sin t

is given by
KA
y (t )  sin  t   
1     
2 2
  2 
2 (29)
where  2 
(30)
   tan  1
2 
1     

The output amplitude from (29) is

KA
Aˆ  (31)
1    
2 2
  2 
2
Output amplitude
Amplitude ratio, AR 
Input amplitude
Aˆ 1
Normalized AR, ARN   (32)
KA 1     
2 2
  2 
2

The maximum value of ARN can be found by

differentiating (32) w.r.t.  and setting the derivative to
zero. Solving for  gives max
1  2 2
 max  for 0 <  < 0.707 (33)

For  > 0.707, there is no maximum.

1
ARN max
 for 0 <  < 0.707 (34)
2 1   2


ANNOUNCEMENT
Extra classes:
28 March, 4 April & 18 April, 2006
8.00 – 10.00 PM
C6, Pocket C

Dynamic Behavior of First-Order and Second-Order Processes: CAB3014 - Chemical Process Dynamics and Control

Uploaded by

Copyright:

Available Formats

Dynamic Behavior of First-Order and Second-Order Processes: CAB3014 - Chemical Process Dynamics and Control

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dynamic Behavior of First-Order and Second-Order Processes: CAB3014 - Chemical Process Dynamics and Control

Uploaded by

Copyright:

Available Formats

Process Dynamics and Control

Dynamic Behavior of First-order

07/25/20 CAB3014 - Chemical Process Dynamics and Control 1

End of this chapter, you should be able to:

1. Determine dynamic behavior of chemical processes which

07/25/20 CAB3014 - Chemical Process Dynamics and Control 2

• Investigate how processes respond to

07/25/20 CAB3014 - Chemical Process Dynamics and Control 3

• For a general first-order transfer function with output Y(s)

a general time-domain solution can be found once the

07/25/20 CAB3014 - Chemical Process Dynamics and Control 4

What are they?

07/25/20 CAB3014 - Chemical Process Dynamics and Control 5

• Inputs that can be manipulated to control

• Inputs that are not manipulated, classified

07/25/20 CAB3014 - Chemical Process Dynamics and Control 6

• The word input refers

07/25/20 CAB3014 - Chemical Process Dynamics and Control 7

• Ramp input (gradual upward or downward

07/25/20 CAB3014 - Chemical Process Dynamics and Control 8

07/25/20 CAB3014 - Chemical Process Dynamics and Control 9

• Impulse input,  (t ) - Dirac delta function

• Random inputs – changes in a complex manner –

07/25/20 CAB3014 - Chemical Process Dynamics and Control 10

• We use the general first-order transfer

07/25/20 CAB3014 - Chemical Process Dynamics and Control 11

• The time-domain response is given by

07/25/20 CAB3014 - Chemical Process Dynamics and Control 12

Expanding into partial fractions, we get

The Heaviside expansion gives

07/25/20 CAB3014 - Chemical Process Dynamics and Control 13

Taking the inverse Laplace transform of eq. (5.6) we get

y (t )  Ka (e  t /   1)  Kat (5.7)

07/25/20 CAB3014 - Chemical Process Dynamics and Control 14

For a first-order process

By partial fraction decomposition

KA 2  KA KA

07/25/20 CAB3014 - Chemical Process Dynamics and Control 15

We have This term dies out for large t

where    tan 1 ( )

For large t, y(t) is also sinusoidal, output sine is attenuated

• Consider a stirred-tank heating system with

Suppose that the metal heating element has a significant

temperature for the heating element).

The unsteady-state energy balance for the tank and the

where m  V and me C e is the product of the mass of the

Q is the thermal equivalent of the instantaneous electrical

Under steady-state conditions, we have

Laplace transform each equation, and after

07/25/20 CAB3014 - Chemical Process Dynamics and Control 20

Eliminate Te(s) from (9) by using (10).

The transfer function relating T  to Q is

and the transfer function relating T  to Ti is

07/25/20 CAB3014 - Chemical Process Dynamics and Control 21

By the principle of superposition

Factoring the characteristic polynomial,

07/25/20 CAB3014 - Chemical Process Dynamics and Control 22

07/25/20 CAB3014 - Chemical Process Dynamics and Control 23

The stirred-tank heating system is used to preheat a