CHE160: Process Dynamics and Control

BS Chemical Engineering

Professor Mary Jane G. Zozobrado

1st Semester, Academic Year 2025-2026

Student Name: Rania-Raihan [Link]

Student ID: 202131059

Machine Problem Number: 1 (Unit 4)

Title of the Machine Problem: Control System

Date Submitted: December 4, 2025

MATLAB Version Used: R2025a

Declaration: I hereby declare that the submission is my own work and that I have not plagiarized or received
unauthorized assistance.

Signature:
Contents
1 Introduction........................................................................................................................................................ 3

2 Problem Statement ............................................................................................................................................. 3

3 Methodology ...................................................................................................................................................... 3

4 MATLAB Code ................................................................................................................................................ 5

5 Results and Analysis .......................................................................................................................................... 5

6 Discussion .......................................................................................................................................................... 7

7 Conclusions........................................................................................................................................................ 8

8 References.......................................................................................................................................................... 9

9 Appendices ........................................................................................................................................................ 9
1 Introduction
Control systems are important in chemical process industries because they maintain the key operating
variables close to the desired set points in spite of disturbances. For a temperature-controlled system like a
stirred-tank heater, a control loop continuously measures the outlet temperature, compares that with a set point,
and regulates the heat input to the system. A typical feedback control loop includes the process, the measuring
element, the controller, and the final control element. These can be represented in a block diagram form, where
each block bears a transfer function describing its dynamic response. The controller would act to minimize the
mismatch between desired and measured temperature and may assume one of several forms, including
proportional, PI, PD, and PID control. Each of these provides a different means of correcting deviations in a
process. Thus, this machine problem deals with the application of those concepts by developing a block diagram
for a stirred-tank heater and simulating its response to step changes. The model will also permit assessment and
understanding of how measurement lag and controller action can affect temperature regulation.

2 Problem Statement
This machine problem aims to construct and simulate the block-diagram representation of the stirred-tank
heater using the transfer-function models derived in Examples 5.2, 8.1, and 8.2. in MATLAB Simulink. The
specific objectives are:

1. Example 8.3 Simulation

For the system given in Example 5.2 of the book, simulate the responses of the actual tank
temperature and the measured temperature (thermocouple output) to a 10 °C step increase in inlet
temperature with no change in heat input.

2. Example 8.4 Simulation

Using a proportional controller with gain Kc = 20, simulate the tank temperature response to a 5 °C
step increase in the set point based on previously derived transfer functions.

The simulations will show how the control system handles disturbances and set-point changes.

3 Methodology
In solving the given problem sets using MATLAB Simulink, the following procedure was followed:

A. Example 8.3

1. The transfer functions needed for the stirred-tank temperature response and thermocouple measurement
1
were already provided in the problem, where the tank is modeled as 5𝑠+1 and the thermocouple behaves
1
as ; therefore, the Simulink model could be constructed directly using these values without
0.33𝑠+1

additional derivations.
 2. MATLAB was opened and the Simulink environment was launched by typing simulink in the
Command Window, which opened the Simulink workspace, allowing the creation of a new model.
3. From the Library Browser, the necessary blocks were inserted into the model workspace, specifically
the Step block to serve as the inlet temperature disturbance, two Transfer Function blocks
for the tank and thermocouple dynamics, a Mux block for combining outputs, and a Scope block
for displaying both temperature responses simultaneously.
4. The Step block parameters were modified to represent a 10°C sudden change in inlet temperature at
time zero by setting the Step Time to 0, the Initial Value to 0, and the Final Value to 10.
5. The first Transfer Function block was edited to define the dynamic behavior of the stirred-tank heater
by setting its numerator to 1 and denominator to [5 1], while the second Transfer Function block was
configured with a numerator of 1 and denominator of [0.33 1] to appropriately represent the
thermocouple’s first-order response.
6. The blocks were then connected. The output of the tank model serves as the actual tank temperature and
is split into two paths: one going directly to the Mux and the other passing through the thermocouple
transfer function so that both the actual and measured temperatures can be displayed for comparison.
7. The simulation stop time was set to 200 seconds to ensure that the full transient response, including the
approach to steady state, would be visible for both temperature signals on the Scope.
8. The model simulation was executed, and once complete, the Scope block was opened to visualize
the time responses of the actual tank temperature and the slower-responding thermocouple output.
Additionally, the To Workspace block sent the same signals into the MATLAB workspace as a
structured variable named [Link], where out is the default simulation output object created by
Simulink and Temps is the user-defined variable name inside the block.
9. The results were analyzed by observing the lag of the indicated temperature relative to the actual tank
temperature, which demonstrates the effect of measurement dynamics in a real feedback control
system.

B. Example 8.4

1. The closed-loop stirred-tank control system in this example uses the same process and measurement
1
dynamics as Example 8.3, namely the tank transfer function 5𝑠+1 and the thermocouple transfer function
1
0.33𝑠+1
. However, a proportional controller is now added with a gain KC = 20, along with a heater gain
of 0.0714. Because all required parameters were provided, the development proceeded directly to model
construction.
2. MATLAB and Simulink were opened in the same manner as in Example 8.3. A new Simulink model
was created, but this time following the feedback control structure shown in Figure 8-17 rather than the
open-loop arrangement previously used.
 3. Similar to the earlier setup, a Step block was inserted to represent the input forcing, but this time as
a set-point change rather than a disturbance. The Step block values were configured to produce a +5°C
command at time zero, with Step Time = 0, Initial Value = 0, and Final Value = 5.
4. A Sum block was placed at the input of the controller and defined to compute the error signal 𝜀 = T’r
− T’m, establishing the negative feedback loop. The output of this block serves as the controller input.
5. Instead of feeding the tank directly with the Step response as in Example 8.3, the control effort first
passes through a Gain block set to KC = 20, which represents the proportional controller. This was
followed by another Gain block set to 0.0714 to model the heater conversion of controller output into
heat input.
6. The configuration of the tank and thermocouple Transfer Function blocks remained the same as in
Example 8.3, but their signal routing changed: the tank output now branches into two paths, one feeding
the display through a Mux and the other passing back through the thermocouple transfer function for
measurement feedback.
7. All blocks were interconnected to replicate the exact closed-loop configuration of Figure 8-17, ensuring
that the measured output continuously influences the controller action through feedback.
8. The simulation stop time was again set to 200 seconds, allowing observation of both transient and steady-
state behavior as the proportional controller attempts to reduce the error between the measured
temperature and the new set-point.
9. The model was executed, and the Scope display was examined to evaluate how effectively the
proportional control adjusted the tank temperature toward the desired 5°C increase, including any steady-
state offset characteristic of P-only control.

4 MATLAB Code
% Machine Problem #1: Control System (Example 8.3 - Step Response of Stirred Tank
and Thermocouple System)
% Author: Magangcong, Rania-Raihan M.
% Date: December 4, 2025
% Run the Simulink Model
% The Simulink file ('Example_8_3_Model.slx') must be located in the same folder
% or added to the MATLAB path before executing the script.
out = sim('Example_8_3_Model');
% Plot the Results
% The output 'Temps' is a timeseries object stored inside the structure 'out'.
% The logged data can be accessed using:
% [Link] = simulation time array
% [Link] = temperature signals array
% Column 1: Actual Tank Temperature
% Column 2: Thermocouple Measured Temperature
plot([Link], [Link])
xlabel('Time (s)') % Label for time axis
ylabel('Temperature Change (°C)') % Output axis label
legend('Indicated Temp Tm prime', 'Actual Temp T prime')

5 Results and Analysis

For Example 8.3, the result is as follows.
 Fig 1a. Simulink Model for the System in Example 8.3

Fig 1b. Comparison of temperature response for measured and actual temperatures in the stirred tank heater.
(i) Complete response; (ii) expanded time scale for short times.

Fig 1c. MATLAB plot of output for stirred-tank heater response

Figures 1b and 1c show the temperature of the stirred-tank outlet versus the thermocouple-indicated
temperature for a 10°C step change in the inlet stream. The true outlet temperature rises quickly, approaching
the full 10°C increase within a few time constants because the tank behaves as a fast first-order process. In
contrast, the thermocouple reading rises much more slowly and never quite equals the real temperature during
the transient period; its indicated value continues to lag by several degrees until steady state is nearly reached.
This clear mismatch highlights that the measurement dynamics are significantly slower than the actual thermal
response of the process.

For Example 8.4, the result is as follows.

Fig 2a. Simulink Model for the System in Example 8.4

Approx. 2.94C

Fig 2b. Response of stirred tank heater control system for proportional control only to a step change in the set
point of 5℃ (Kc= 20)

Figure 2b presents the response of the closed-loop system to a 5°C step change in set point for
proportional-only control with Kc =20. The real tank temperature increases gradually and smoothly, reflecting
that the system is stable and overdamped. However, the steady-state value reaches only about 2.94°C instead of
the desired 5°C, showing that a proportional-only controller cannot drive the error to zero.

6 Discussion
The results illustrate the significant impact of measurement dynamics on closed-loop control
performance. In Figures 1b and 1c, the thermocouple’s slow response creates a lag between the true outlet
temperature and the measured signal. This behavior illustrates an important principle of feedback control: the
controller acts on the measured signal, not on the true process variable. When the measurement element is
sluggish, the controller receives outdated information and may respond as if the tank is still far from its setpoint.
This delayed perception can make the controller overcorrect, leading to slower stabilization and potentially to
overshoot or oscillations if the controller gain is not chosen carefully. Thus, Example 8.3 emphasizes that closed-
loop performance depends on both the process and measurement dynamics. The insight becomes especially
important as we transition to Example 8.4, where a proportional controller is introduced. Because the controller
will rely on the delayed thermocouple output, the interaction of these elements must be accounted for to achieve
a stable and well-tuned temperature-control system

Figure 2b shows the effect of this measurement lag when proportional-only control is applied. The
controller will dampen the process and eliminate oscillations, but the slow thermocouple will reduce the
corrective action prematurely. The result is an overdamped, stable response, where the actual tank temperature
increases smoothly and slowly. Significantly, the system converges to a value well below the set point of about
2.94°C versus the target of 5°C. The reasons for this steady-state offset are that proportional control cannot drive
the process error to zero in a nonintegrating process, and a sustained error is required to yield an appropriate
control action to maintain heating. This situation is further aggravated by sluggish thermocouple dynamics,
which make the controller think that the temperature is closer to the set point than it actually is, thereby reducing
the corrective effort prematurely. Example 8.4 thus illustrates an important lesson from control design: even
though a controller stabilizes the process, use of proportional action alone in conjunction with measurement lag
results in sluggish response with sustained error. This provides the motivation for integral action as will be
developed in the next section to force the error to zero and improve the accuracy of the closed-loop temperature
control.

7 Conclusions
The MATLAB Simulink simulations for the stirred-tank heater system illustrate some important
principles of feedback control in regulating temperature. The slow response of the thermocouple in Example 8.3
resulted in a lagged measured signal with respect to the tank temperature itself which is a consequence of
measurement dynamics in controller perception. Example 8.4 involved the stabilization of the process by a
proportional controller for a 5°C setpoint change. The resultant overdamped response exhibited a steady-state
error of about 2.94°C. The explanation is that pure proportional action cannot eliminate the offset in a
nonintegrating system, and the lag in measurement reduced corrective action further.

Overall, results highlight that both process and measurement dynamics are important for closed-loop
performance. They also illustrate why proportional control might not be satisfactory for accurate temperature
control and justify the implementation of integral action in order to eliminate steady-state error and achieve
greater closed-loop accuracy. The results again underscore that controller design and measurement
characteristics are equally important for feedback control in chemical processes.
8 References
[1] D. R. Coughanowr and S. LeBlanc, Process Systems Analysis and Control, 3rd ed. New York, NY, USA:
McGraw-Hill, 2009

9 Appendices

The reference system of this machine problem is the system described in Example 5.2. This problem is as
follows:

𝑇′(𝑠) 1
Where the derived transfer function is: =
𝑇𝑖 ′(𝑠) 5𝑠+1

Note also that Example 8.3 is a continuation of Example 8.2. The transfer function derived in this problem is as
follows:

And the entire system is also represented by a block flow diagram:

Figure A1. Block diagram for stirred-tank heater and measuring element.

For Example 8.4, the following transforms are used:

The system can also be represented by block flow diagrams:

Figure A2. Block diagram of proportional controller.

Figure A3. Block flow diagram of the control system as given in Example 8.4.