Lecture #9-10

2024. 04. 09(Tue.)

Hyun-Seok Cho, Ph.D.

Chemical and Biomolecular Engineering

Sogang University

CBE3006-Spring 2024, _040924_K303

A sudden change in a process variable can be approximated by a
step change of magnitude, M:

0 𝑡𝑡 < 0
𝑈𝑈𝑠𝑠 = �
𝑀𝑀 𝑡𝑡 ≥ 0

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, S(t).

𝑀𝑀
𝑈𝑈 𝑠𝑠 =
𝑠𝑠

2
Example of a step change:

A reactor feedstock is suddenly switched from one supply to

another, causing sudden changes in feed concentration, flow, etc.

The heat input to the stirred-tank heating system is suddenly

changed from 8000 to 10,000 kcal/hr by changing the electrical
signal to the heater.

Thus,

𝑄𝑄 𝑡𝑡 = 8,000 + 2,000 𝑆𝑆 𝑡𝑡 , 𝑆𝑆 𝑡𝑡 = unit step

𝑄𝑄′ 𝑡𝑡 = 𝑄𝑄 − 𝑄𝑄̄ = 2,000 𝑆𝑆 𝑡𝑡 , ̄𝑄𝑄 = 8,000 kcal/hr

3
• 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:

𝑎𝑎
𝑈𝑈𝑅𝑅 𝑡𝑡 = � 0 𝑡𝑡 < 0 𝑈𝑈 𝑠𝑠 = 2
a∗t 𝑡𝑡 ≥ 0 𝑠𝑠

Examples of ramp changes:

1. Ramp a setpoint to a new value. (Why not make a step

change?)
2. Feed composition, heat exchanger fouling, catalyst activity,
ambient temperature.
4
• It represents a brief, sudden change in a process variable:

0 for 𝑡𝑡 < 0
𝑈𝑈𝑅𝑅𝑅𝑅 𝑡𝑡 = �ℎ for 0 ≤ 𝑡𝑡 < 𝑡𝑡𝑤𝑤
𝑡𝑡𝑤𝑤
0 for 𝑡𝑡 ≥ 𝑡𝑡𝑤𝑤

ℎ
𝑈𝑈 𝑠𝑠 = [1 − 𝑒𝑒𝑒𝑒𝑒𝑒 −𝑡𝑡𝑡𝑡𝑡𝑡 ]
𝑠𝑠
Examples:

1. Reactor feed is shut off for one hour.

2. The fuel gas supply to a furnace is briefly interrupted.
5
6
7
• Processes are also subject to periodic, or cyclic, disturbances.
They can be approximated by a sinusoidal disturbance:
where: A = amplitude, ω = angular frequency

0 for 𝑡𝑡 < 0
𝑈𝑈sin 𝑡𝑡 �
𝐴𝐴 sin 𝜔𝜔𝜔𝜔 for 𝑡𝑡 ≥ 0
𝐴𝐴ω
𝑈𝑈 𝑠𝑠 = 2
𝑠𝑠 + ω2
Examples:

1. 24 hour variations in cooling water temperature.

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

8
• It represents a short, transient disturbance. U I (t ) = δ (t ).

Examples: 𝑈𝑈 𝑠𝑠 = 1

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,

𝑌𝑌 𝑠𝑠 = 𝐺𝐺 𝑠𝑠 𝑈𝑈 𝑠𝑠

9
The corresponding time domain express is:

𝑡𝑡
𝑦𝑦 𝑡𝑡 = � 𝑔𝑔 𝑡𝑡 − τ 𝑢𝑢 τ 𝑑𝑑τ
0

𝑔𝑔 𝑡𝑡 = 𝔏𝔏−1 𝐺𝐺 𝑠𝑠

Suppose u ( t ) = δ ( t ) . Then it can be shown that:

𝑦𝑦 𝑡𝑡 = 𝑔𝑔 𝑡𝑡

Consequently, g(t) is called the “impulse response function”.

10
For a general first-order transfer function with output Y(s) and input U(s),

𝑌𝑌 𝑠𝑠 𝐾𝐾 𝐾𝐾 = steady−state gain
=
𝑈𝑈 𝑠𝑠 𝜏𝜏𝜏𝜏 + 1 τ = time constant

Consider the response of this system to a step of magnitude, M:

𝑀𝑀
𝑈𝑈 𝑡𝑡 = 𝑀𝑀 for 𝑡𝑡 ≥ 0 ⇒ 𝑈𝑈 𝑠𝑠 =
𝑠𝑠

𝐾𝐾𝐾𝐾
𝑌𝑌 𝑠𝑠 =
𝑠𝑠 τ𝑠𝑠 + 1
11
Step response

𝐾𝐾𝐾𝐾
𝑌𝑌 𝑠𝑠 = 𝑦𝑦 𝑡𝑡 = 𝐾𝐾𝐾𝐾 1 − 𝑒𝑒 −𝑡𝑡/τ
𝑠𝑠 τ𝑠𝑠 + 1
𝑦𝑦∞ = steady-state value of y(t).
y∞ = KM .
y
y∞
t ___
0 0
0.632
Note: Large τ means a slow response.
0.865
0.950
0.982
0.993
12
Ramp response

𝑎𝑎 𝐾𝐾𝐾𝐾
𝑌𝑌 𝑠𝑠 = 2
𝑠𝑠 τ𝑠𝑠 + 1

y(t) =K∗𝑎𝑎∗t − K∗τ∗𝑎𝑎+ K∗τ∗𝑎𝑎∗exp(−t/τ)

t >>τ, y(t) =K𝑎𝑎(t−τ)

13
Sinusoidal response

𝐴𝐴ω 𝐾𝐾
𝑌𝑌 𝑠𝑠 = 2
𝑠𝑠 + ω2 τ𝑠𝑠 + 1

𝐾𝐾𝐾𝐾
𝑦𝑦 𝑡𝑡 = (ωτ exp(-t/τ) -ωτ cos ωt + sin ωt)
1+ω2τ2

𝐾𝐾𝐾𝐾ωτ 𝐾𝐾𝐾𝐾
𝑦𝑦 𝑡𝑡 = exp(-t/τ) + sin (ωt+ф) ф = -tan-1 (ωt)
1+ω2τ2 (1+ω τ )2 2

ω = frequency in radian[radians/time]
ω = 2π ∗ cycle frequency cycles
τ=[ ]
s
the period = 2π/ω
14
• Not all processes have a steady-state gain. For example, an
“integrating process” or “integrator” has the transfer function:

𝑌𝑌 𝑠𝑠 𝐾𝐾
= 𝐾𝐾 = constant
𝑈𝑈 𝑠𝑠 𝑠𝑠

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

𝐾𝐾𝐾𝐾
𝑌𝑌 𝑠𝑠 = 2 ⇒ 𝑦𝑦 𝑡𝑡 = 𝐾𝐾𝐾𝐾𝐾𝐾
𝑠𝑠

Thus, y(t) is unbounded and a new steady-state value does not exist.

15
Consider a liquid storage tank with a pump on the exit line:

- Assume:
1. Constant cross-sectional area, A.
2. q ≠ f ( h )
𝑑𝑑𝑑
- Mass balance: 𝐴𝐴 = 𝑞𝑞𝑖𝑖 − 𝑞𝑞 (1) ⇒ 0 = 𝑞𝑞̄ 𝑖𝑖 − 𝑞𝑞̄ (2)
𝑑𝑑𝑑𝑑

- Eq. (1) – Eq. (2), take L, assume steady state initially,

1
′
𝐻𝐻 𝑠𝑠 = 𝑄𝑄𝑖𝑖′ 𝑠𝑠 − 𝑄𝑄′ 𝑠𝑠
𝐴𝐴𝐴𝐴 𝐻𝐻′ 𝑠𝑠 1
- For Q′ ( s ) = 0 (constant q), ′ =
𝑄𝑄𝑖𝑖 𝑠𝑠 𝐴𝐴𝐴𝐴
16
For a second-order transfer function with output Y(s) and input U(s),

𝑌𝑌 𝑠𝑠 𝐾𝐾
= 2 2
𝑈𝑈 𝑠𝑠 τ 𝑠𝑠 + 2ζτ𝑠𝑠 + 1
𝐾𝐾 = steady−state gain
τ = "time constant" [=] time
ζ = damping 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 (dimensionless)
𝜔𝜔𝑛𝑛 = natural frequency of the system
1
𝑌𝑌 𝑠𝑠 𝐾𝐾𝜔𝜔𝑛𝑛2 =
= 2 τ
𝑈𝑈 𝑠𝑠 𝑠𝑠 + 2ζ𝜔𝜔𝑛𝑛 𝑠𝑠 + 𝜔𝜔𝑛𝑛2

𝑠𝑠 = −ζ𝜔𝜔𝑛𝑛 ± 𝑗𝑗𝜔𝜔𝑛𝑛 (1 − ζ2)

17
Damping Type of Response Roots of Charact.
Coefficient Polynomial
Overdamped Real and unequal
ζ>1

ζ=1 Critically damped Real and repeated

Underdamped Complex conjugates

0≤ζ<1
𝑠𝑠 = −ζ𝜔𝜔𝑛𝑛 ± 𝑗𝑗𝜔𝜔𝑛𝑛 (1 − ζ2)

• Note: The characteristic polynomial is the denominator of the

transfer function: 2 2
τ s + 2ζτs + 1
𝑠𝑠 = −ζ𝜔𝜔𝑛𝑛 ± 𝜔𝜔𝑛𝑛 (ζ2 − 1)

• What about ζ < 0 ? It results in an unstable system

18
Underdamped

19
Critically damped and overdamped

20
1. Responses exhibiting oscillation and overshoot (y/KM > 1) are
obtained only for values of ζ less than one.
2. Large values of ζ yield a sluggish (slow) response.
3. The fastest response without overshoot is obtained for the
critically damped case ( ζ = 1) .

21
1. Rise Time: tr is the time the process output
takes to first reach the new steady-state value.
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% 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.
22
A heater for a semiconductor wafer has first-order dynamics, that is, the transfer function
relating changes in temperature T to changes in the heater input power level P is

𝑇𝑇 𝑠𝑠 𝐾𝐾
=
𝑃𝑃 𝑠𝑠 τ𝑠𝑠 + 1
where K has unit [oC/kW] and τ has units [min]
The process is at a steady state when an engineer changes the power input stepwise from 1
to 2 kW. She notes the following:
1) The process temperature initially is 100oC.
2) 4 min after changing the power input, the temperature is 400oC.
3) 30 min later the temperature is 500oC.

Q. What are K and τ in the process transfer function?

23
A thermocouple has the following characteristics when it is immersed in a stirred bath:

Mass of thermocouple = 1 g
Heat capacity of thermocouple = 0.25 cal/g oC
Heat transfer coefficient = 20 cal/cm2 h oC (for thermocouple and bath)
Surface area of thermocouple = 3 cm2

(a) Derive a transfer function model for the thermocouple relating the change in its
indicated output T to the change in the temperature of its surrounding Ts assuming
uniform temperature, no conduction in the leads, constant physical properties, and
conversion of millivolt-level output directly to a oC reading by a very fast sensor
(b) If the thermocouple is initially out of the bath and at room temperature (23 oC), what is
the maximum temperature that it will register if it is suddenly plunged into the bath(80
oC) and held there for 20 sec?

24
A process is described by the following transfer function

𝑌𝑌 𝑠𝑠 𝐾𝐾
=
𝑈𝑈 𝑠𝑠 s(τ𝑠𝑠 + 1)

Thus, it exhibits characteristics of both first-order and integrating processes.

How could you utilize a step change in the input of magnitude M to find quickly the two
parameters K and τ ?

25
26