Temperature and Flow Measurement

MECH3316 - A

Alfred Juliant Susanto

2022360013
Submission date: 9 May 2025

Faculty of Engineering and Technology

Sampoerna University
Spring 2025

Chapter 1: Introduction

1.1 Overview
Instrumentation plays a critical role in engineering systems to monitor, measure, and regulate
vital physical variables like temperature and flow rate. These variables directly influence the
safety, efficiency, and performance of processes in industrial and laboratory environments.
Temperature measurement involves quantifying thermal conditions using sensors that detect
physical changes like resistance, voltage, or material expansion [1]. Meanwhile, flow
measurement refers to determining the movement of fluids through pipes, often through
volume or mass flow rates [2].
 In this laboratory experiment, two primary objectives were explored: evaluating the
performance of various temperature sensors using the Hampden Model H-6806 Temperature
Measurement Demonstrator and studying flow rate control using a paddle wheel sensor and
PID controller on the Hampden Model H-ICS-FL-TT-CDL Process Trainer. The aim was to
understand how measurement principles translate into real-world system monitoring and to
assess the effectiveness and stability of control loops in flow systems.

1.2 Background
Precise measurement and control of physical parameters such as temperature and fluid
flow are fundamental in engineering systems. These parameters are crucial for maintaining
process stability, product quality, and equipment safety in industries ranging from
manufacturing and HVAC to food processing and chemical plants. The present experiment
focuses on two measurement modules: temperature sensing and flow measurement with
control feedback.
Temperature measurement is performed using a range of sensors that operate based on
different physical principles:
 Thermocouples function by generating a voltage through the Seebeck effect, where
two dissimilar metals create a measurable electromotive force due to a temperature
gradient. These sensors are robust and suitable for a wide range of temperatures.
 RTDs (Resistance Temperature Detectors) use the predictable change in electrical
resistance of metals like platinum with temperature. RTDs are valued for their high
accuracy and stability over time.
 Thermistors are temperature-sensitive resistors made from semiconductor materials.
Their resistance changes rapidly with temperature, making them suitable for
applications requiring high sensitivity.
 Standard liquid-in-glass thermometers serve as reference instruments due to their
long-standing reliability and direct temperature reading.

1.3 Objectives
 To understand how to use different types of temperature sensor.
 To identify the advantages and limitations of each temperature sensor.
 To understand how to do flow measurement.
 To determine the errors.

Chapter 2: Methodology

2.1 Apparatus and Test Setups

 The temperature measurement module was conducted using the Hampden Model H-
6806 Temperature Measurement Demonstrator, a versatile apparatus designed to simulate
real-world temperature sensing applications as shown in Figure 2.1. This unit includes
features to control air temperature, house sensors, and compare outputs in real-time. The
following sensors were utilized in the experiment:
 RTD (Resistance Temperature Detector): An RTD measures temperature based on the
predictable increase in electrical resistance of certain metals, typically platinum, as
temperature rises [3]. The resistance is approximately linear with temperature,
allowing for accurate and repeatable measurements in industrial applications [4]. The
relationship between resistance and temperature in an RTD is given by:
RT =R0 (1+∝ ∙ ∆ T ) (Formula 1)
 Thermistor: Thermistors are semiconductor-based temperature sensors whose
resistance varies non-linearly with temperature. They are highly sensitive within
limited temperature ranges and are ideal for detecting small thermal variations [5].
The resistance of a thermistor as a function of temperature can be expressed using the
following equation:
1 1
β( − )
R(T )=R0 ∙ e T T0 (Formula 2)
 Liquid Thermometer: A liquid thermometer utilizes the thermal expansion of a fluid,
commonly mercury or alcohol, contained within a sealed glass capillary tube. As
temperature increases, the liquid expands and rises, as temperature decreases, the
liquid contracts and falls. The height of the liquid column corresponds to the
measured temperature, making it a straightforward and widely used method for
thermal indication [6].
 Thermocouple (analogue and digital): Thermocouple is a temperature-sensing device
made by joining two dissimilar metal wires at one end, forming what is known as the
hot junction. When there is a temperature difference between the hot junction and the
cold junctions (the open ends of the wires), a voltage is generated due to the Seebeck
effect. This voltage is directly related to the temperature difference and can be used to
determine the unknown temperature [7]. The thermoelectric voltage is governed by:
V =α ∙ ∆ T (Formula 3)
 Flow measurement plays a vital role in managing fluid systems across a wide range of
applications, including chemical manufacturing, water supply systems, and thermal energy
regulation. It refers to determining the volume of fluid passing through a system per unit
time, commonly measured in litres per minute or gallons per minute [8]. In this practicum,
flow rate was monitored using a paddle wheel flowmeter, a device that detects fluid velocity
by translating the rotational speed of a small internal paddle into an electrical signal. This
rotation generates a 4–20 mA analogue signal, which represents the instantaneous flow rate in
the system.
To regulate this flow, the system utilized a Proportional–Integral–Derivative (PID) controller,
a widely adopted feedback mechanism in industrial process control. The controller
continuously compares the process variable (PV), the current flow rate measured by the
sensor, with a user-specified set point (SP) [9]. Based on this comparison, the controller
adjusts an electrically actuated valve to either increase or decrease the fluid flow, ensuring
that the system approaches and maintains the target value.
The PID controller’s behaviour is governed by a composite control equation that incorporates
three elements: proportional, integral, and derivative responses. These components are
mathematically represented in the standard PID formulation:
t
de (t)
u ( t )=K p e ( t ) + K i∫ e ( τ ) dτ + K d (Formula 4)
0 d (t)

where e(t) denotes the error signal, defined as the difference between the set point and the
process variable at time t, and Kp, Ki, and Kd represent the respective gain constants for each
control term. Although this experiment did not involve direct calculation of the controller
output, familiarity with this equation is essential for understanding how tuning each
parameter influences system performance. To evaluate the control system’s precision, the
error can also be determined by calculating the deviation between the set point and the
average value of the measured process variable, given by:
Error=SP−PV (Formula 5)
These foundational concepts are essential for interpreting the experimental data presented in
subsequent sections, where the effects of adjusting the PID parameters and sources of
measurement inaccuracy are analysed.

2.2 Experimental Procedure

 In Part 1, known weights as shown in Figure 2.1 were added incrementally to the calibrator
piston, starting from the base value as shown in Figure 2.2. After each new weight was
applied, the system was given a few moments to stabilize. The pressure reading shown on the
Bourdon gauge was then recorded as demonstrated in Figure 2.3. This process was repeated
until the full range of weights was used. After reaching the maximum load, weights were
gradually removed one at a time in the same order, and the corresponding pressure readings
were taken again during the decreasing phase. This allowed identification of any hysteresis,
where the gauge may have shown different readings for the same pressure depending on
whether the pressure was increasing or decreasing.

Figure 2.1 Weighing Plates Measurement

Figure 2.2 Dead Weight Calibrator Set-up

 Figure 2.3 Bourdon Gauge
In Part 2, the pressure regulator was used to increase the input air pressure in uniform steps.
At each step, the displayed pressure readings from the strain gauge, semiconductor beam,
piezo-resistive sensor from the Pressure Measurement Bench as shown in Figure 2.4, and the
analog Bourdon gauge were recorded. After reaching the upper limit of the compressor’s
range, the process was repeated in reverse by decreasing the pressure step-by-step. This
ensured that both the increasing and decreasing pressure cycles were covered. The recorded
readings were then used to compare the accuracy, linearity, and hysteresis behavior of each
transducer.

`
Figure 2.4 Pressure Measurement Bench
Chapter 3: Results & Discussion

3.1 Results of Part 1

Table 3.1 Measurements of Temperature Results with Liquid Thermometer

Thermometer (°C)

No Specimens 0 10 20 30 40 50 60

Room
1 Temperature 22 23 23 23 23 23 23

2 Water 21 21 22 22 23 23 23

3 Hot Water 40 52 59 70 76 81 83

4 Ice Cube 20 18 14 11 10 9 8
 The Bourdon pressure gauge showed relatively low error margins across the tested pressure
range. At 200 psi, the error was minimal (1.00% during loading), confirming a high degree of
accuracy in the mid-range. However, as pressure increased to 400 psi, the error grew to
4.50% during unloading, indicating the influence of mechanical limitations such as friction
and internal hysteresis. The fact that the loading and unloading values differed slightly also
illustrates this hysteresis, though it remains within acceptable tolerance for many industrial
applications. The data suggest that while the Bourdon gauge performs well across typical
pressure ranges, its accuracy slightly declines at higher extremes.

Figure 3.1 Actual Pressure against Measured Pressure (Gauge Reading)

The plotted calibration curve of nominal pressure against gauge readings shows a linear
relationship for both loading and unloading cycles. The two lines follow similar trends but
are not perfectly overlapping, confirming the presence of hysteresis. The largest deviation
between the curves appears at 400 psi, indicating that the gauge may slightly under-read
during unloading. The overall linearity of the curves confirms that the Bourdon gauge
behaves predictably under increasing and decreasing pressure cycles, with only minor lag at
higher pressures.
Table 3.2 Measurements of Temperature Results with Analog Thermocouple
 Thermocouple Analog (°C)

No Specimens 0 10 20 30 40 50 60

Room
1 Temperature 18 18 18 18 18 19 19

2 Water 19 20 20 20 20 20 20

3 Hot Water 40 78 90 94 95 96 96

4 Ice Cube 20 14 10 4 1 0 -2
 Figure 3.2 Percentage Error against Measured Pressure (Gauge Reading)
This graph highlights how measurement error changes with pressure. Errors were
smallest at 200 psi and increased slightly at higher pressures. The increase in error during
unloading suggests that internal mechanical resistance or friction within the Bourdon gauge
may affect needle return accuracy. Despite the increase, all errors remained under 5%, which
is generally acceptable for analog instruments in industrial applications.

Table 3.3 Measurements of Temperature Results with Digital Thermocouple

Thermocouple Digital (°C)

No Specimens 0 10 20 30 40 50 60

Room
1 Temperature 22.8 22.9 22.9 22.8 22.9 22.9 22.9

2 Water 23.1 23.7 23.7 23.7 23.7 23.7 23.7

3 Hot Water 63.5 104 104.1 104 104.2 104.2 104.3

4 Ice Cube 19.3 1.6 0 -0.7 -0.8 -1 -1.3

 Table 3.4 Measurements of Temperature Results with RTD

RTD (°C)

No Specimens 0 10 20 30 40 50 60

Room
1 Temperature 22 22.2 22.2 22.1 22.2 22.3 22.3

2 Water 23.1 23.3 22.5 23.3 23.5 23.5 23.5

3 Hot Water 60.15 100.7 103.7 103 103.2 103.4 103.4

4 Ice Cube 19 17 17.1 12.2 10.1 7.9 5.8

Table 3.5 Measurements of Temperature Results with Thermistors

Thermistors (°C)

No Specimens 0 10 20 30 40 50 60

Room
1 Temperature 22.6 22.8 22.9 22.9 22.9 23 23

2 Water 23.1 23.3 22.5 23.5 23.5 23.7 23.5

3 Hot Water 60.7 100.7 103.7 103 103.5 103.3 103

4 Ice Cube 19 17.3 17.1 12 10 7 5.5

 3.2 Results of Part 2
Table 3.6 Process Variable (PV) Data for Flow Control at Various P, I, and D Settings
 Process
Set Point Time (s) Variable P I D Remarks

12 10 81.5 34

20 35.6 21

30 17.3 17
438 2 0
40 14.1 16.5

50 12 16.3

60 12.6 16.1

10 86.23 43.5

20 40.75 23.4

30 18.55 18.2
438 2 1
40 15.3 16.6

50 12.64 16.1

60 11.74 16

10 67.43 33.6

20 39.78 23.7

30 23.54 19.3
438 3 0
40 16.43 18

50 12.28 17.4

60 12.37 17.1

10 26.55 438 3 1 20

20 17.7 17.9

30 15.15 16.9

40 13.84 16.4

50 12.32 16
 60 11.44 16.1

10 26.51 20.1

20 18.83 18.6

30 17.86 17.5
438 4 0
40 15.73 16.7

50 13.81 16.2

60 12.64 16

10 26.20 20.3

20 18.74 18.7

30 17.44 17.7
438 4 1
40 15.42 16.9

50 14.38 16.5

60 12.34 16.1

Table 3.7 Flow Control Mean PV, Standard Deviation, and Error at Different PID Settings

Standard
Set Deviation Error (SP-
Point P I D Mean PV PV PV)

2 0 40.25667 32.05968 -28.2567

2 1 16.55667 6.930574 -4.55667

3 0 16.30167 4.598396 -4.30167

12 438
3 1 15.11 4.148916 -3.11

4 0 16.265 4.137402 -4.265

4 1 14.68667 2.521254 -2.68667

 The values in Table 3.2 highlight the upper pressure limits for each sensor, reflecting their
design and intended applications. Both the strain-gauge diaphragm sensor and the
semiconductor-on-bending beam sensor have the highest operating range at approximately
33.5 bar (3346–3347 kPa), making them suitable for high-pressure systems or industrial
applications requiring a broad measurement span. In contrast, the piezo-resistive sensor has
a significantly lower limit of 6.89 bar (689 kPa), indicating its usage is more suited for lower
pressure ranges or sensitive measurements where high resolution is prioritized over range.

3.2 Results of Part 2

Table 3.8 Process Variable (PV) Data for Level Control at Various P, I, and D Settings
 Time
Set Point (s) Process Variable P I D Remarks

10 10 9.2 34.1

20 12.23 32.5

30 13.6 31
80 23 0
40 14.47 30

50 14.32 28.9

60 13.99 28.4

10 9.14 34.2

20 11.9 32.2

30 13.34 31.1
80 23 1
40 14.1 29.8

50 14.17 28.9

60 13.87 28.5

10 9.04 32.8

20 11.31 31.6

30 12.73 30.6
80 24 0
40 13.45 29.5

50 13.55 28.8

60 13.23 28.2

10 8.96 80 24 1 33

20 11.28 31.6

30 12.84 30.7

40 13.49 29.6
50 13.57 28.9

60 13.25 28.5

10 9.46 33.3

20 11.9 32.3

30 13.4 30.9
80 25 0
40 14.1 29.9

50 14.3 28.9

60 13.78 28.4

10 10.3 37

20 14 34

30 15.73 32.3
80 25 1
40 16.83 30.6

50 16.87 29.4

60 16.17 28.4
Table 3.9 Level Control Mean PV, Standard Deviation, and Error at Different PID Settings

Set Standard
Point P I D Mean PV Deviation PV Error (SP-PV)

8.906666
23 0 7 1.279683 1.093333

11.53666
23 1 7 1.697917 -1.53667

24 0 11.72 1.591676 -1.72

10 80
11.31666
24 1 7 1.483775 -1.31667

11.15333
25 0 3 1.467193 -1.15333

11.88333
25 1 3 1.665219 -1.88333
best accuracy. The maximum error remained under 5%, and its readings closely matched the
pressure source values throughout both increasing and decreasing cycles. This consistent
response confirms its suitability for applications requiring high precision.
In contrast, the semiconductor-on-bending beam sensor (P2) demonstrated the highest
percentage errors, particularly at low and mid-range pressures, with values exceeding 20%.
These large deviations may be due to poor low-end sensitivity or calibration drift. However,
its values became more stable at higher pressures.
The piezo-resistive sensor (P3) exhibited similar error behavior to P2, with high initial error
(26.03% at 10 psi) that gradually decreased with higher input pressure. These sensors are
often more nonlinear at lower pressures, which was evident in this case.
The Bourdon gauge (P4), as expected, did not respond to low pressure (no deflection at 10
psi), but aligned well between 30–50 psi, showing acceptable error values (as low as 3.31%).
However, its inability to detect lower pressures highlights its mechanical limitations when
compared to digital alternatives.

Figure 3.3 Relation between P1 and Compressor (Pressure Source)

Figure 3.4 Relation between P2 and Compressor (Pressure Source)

Figure 3.5 Relation between P3 and Compressor (Pressure Source)

Figure 3.6 Relation between P41 and Compressor (Pressure Source)

 These graphs compares the pressure readings of the strain gauge (P1) shown in Figure 3.3,
semiconductor (P2) shown in Figure 3.4, piezo-resistive (P3) shown in Figure 3.5, and
Bourdon gauge (P4) shown in Figure 3.6 against the nominal pressure. The strain gauge
sensor (P1) consistently followed the reference pressure, making it the most accurate. P2 and
P3 showed significant overshoot at lower pressures, especially at 10 and 20 psi. The Bourdon
gauge underperformed at lower pressures, showing no response at 10 psi and underestimating
at 20 psi. However, it aligned more closely at higher pressures (30–50 psi). This comparison
emphasizes the need for selecting appropriate sensors based on expected pressure range and
required accuracy.
The error trend clearly shows that P1 had the lowest and most stable error percentages,
typically under 5%. P2 and P3 exhibited higher errors, especially at low pressures, with some
values exceeding 25%. This trend confirms their reduced accuracy near the lower operational
limit. The Bourdon gauge displayed negligible error at high pressures but was less responsive
at the low end.

Chapter 4: Conclusion
This experiment successfully demonstrated the process of calibrating a mechanical Bourdon
pressure gauge using a dead weight calibrator and comparing its performance to a variety of
digital pressure transducers. In Part 1, the Bourdon gauge exhibited low error margins—
ranging from 1.00% to 4.50%—with slight hysteresis observed during unloading cycles.
These results confirm that the gauge maintains acceptable accuracy for typical industrial
applications, especially in the mid-pressure range. However, mechanical limitations such as
internal friction and hysteresis lead to reduced precision at higher pressures.
In Part 2, digital sensors were evaluated for their accuracy and consistency. The strain gauge
diaphragm sensor emerged as the most accurate and stable device, with minimal error and
excellent linearity throughout the tested range. The semiconductor-on-bending beam and
piezo-resistive sensors showed significantly higher errors, especially at lower pressures,
suggesting limitations in sensitivity or calibration drift. The Bourdon gauge, while less
responsive at lower pressures, performed reliably at mid-to-high pressure levels.
Overall, this experiment highlights the importance of selecting appropriate pressure
measurement devices based on the operating range, required accuracy, and dynamic
conditions. It also reinforces the value of calibration procedures in ensuring the reliability of
pressure readings, particularly when instruments are used in safety-critical or high-
performance applications.
Chapter 5: References
[1] J. W. Valvano, “Temperature measurements,” in Advances in Heat Transfer, vol. 22,
Elsevier, 1992, pp. 359–436.
[2] S. Basu, Plant Flow Measurement and Control Handbook: Fluid, Solid, Slurry and
Multiphase Flow, 1st ed. London, U.K.: Academic Press, 2019. [Link]
0-12-812437-6.00001-9
[3] J. Li, H. Pei, O. Kochan, C. Wang, R. Kochan, and A. Ivanyshyn, "Method for
Correcting Error Due to Self-Heating of Resistance Temperature Detectors Suitable for
Metrology in Industry 4.0," Sensors, vol. 24, no. 24, p. 7991, Dec. 2024.
[Link]
[4] M. Cappelli, "Fundamentals of Analog I&C Systems," in Instrumentation and Control
Systems for Nuclear Power Plants, 1st ed., Woodhead Publishing Series in Energy, 2023, pp.
67–115. [Link]
[5] D. Camuffo, “Measuring Instruments and Their Development,” in Microclimate for
Cultural Heritage: Measurement, Risk Assessment, Conservation, Restoration, and
Maintenance of Indoor and Outdoor Monuments, 3rd ed., Amsterdam, Netherlands: Elsevier,
2019, pp. 383–429. [Link]
[6] H. D. Goodfellow and Y. Wang, “Experimental Techniques,” in Industrial Ventilation
Design Guidebook: Volume 2: Engineering Design and Applications, 2nd ed., Academic
Press, 2021, pp. 185–277. [Link]
[7] X. Zhu, Y. Yu, and F. Li, "A review on thermoelectric energy harvesting from asphalt
pavement: Configuration, performance and future," Construction and Building Materials, vol.
228, p. 116818, Dec. 2019. [Link]
[8] Food Chemistry, Elsevier. [Link]
[9] M. Kutz, Handbook of Farm, Dairy, and Food Machinery, 1st ed. Norwich, NY, USA:
William Andrew Publishing, 2007.
[Link]
machinery