You are on page 1of 5

Proceedings of the 2nd MAE paper writing

MAE241
Spring 2011, Coral Gables, Florida, USA

TVT2011-241
TEMPERATURE MEASUREMENT USING A RESISTANCE TEMPERATURE
DETECTOR AND THERMOCOUPLE
Thien Van Tran
MAE 241
Coral Gables, Florida, USA

ABSTRACT
A resistance temperature detector and thermocouple
were used to measure the time constant as a result of
step changes in temperature. The RTD produced 3.1295
1.6195s for the step up in temperature, and 4.1665
3.56s for the step down. The thermocouple produced
0.5620 0.1788s for the step up and 0.4228 2.1857s
for the step down. For the RTD, time of rise from 10% to
90% was 6.925 1.955 seconds, while the fall time was
9.1546 7.802 seconds. For the thermocouple, the rise
time was 1.235 0.393 seconds while the fall time was
0.929 4.802 seconds. All calculations were associated
with a 95% confidence interval.

NOMENCLATURE
T
Temperature
Temperature reached at infinity
Initial temperature
s
Seconds
t
Time
Uncertainty associated with precision

INTRODUCTION
Resistance Temperature Detectors (RTDs) and
Thermocouples are both first order instruments for
measuring temperature. The RTD measures the change
in temperature by detecting changes in resistance as a
small current is passed through the resistor element. This
takes advantage of the accurate direct relationship
between temperature and resistance. The RTD consists
of a length of fine coiled wire wrapped around a core of
glass or ceramic. The RTD is made of a single metal,
whose resistance at temperatures is documented and
standardized.
The thermocouple measures voltage changes
due to temperature changes [1]. It is made up of two
dissimilar metals, joined together at a single end. They
appear in the form of a wire. The RTD is used for
accurate readings over narrow temperature spans, while
the thermocouple is used for reading extreme ranges [2].
Therefore the RTD will have a much longer response
than the thermocouple. Using sudden extreme
temperature differences, the time constant () for rise and
fall of temperature was determined for each sensor.

Greek Symbols

Students t-distribution value


1 confidence interval

Standard deviation
Number of samples

Rise time constant


1 confidence interval

EXPERIMENT DESIGN AND PROCEDURE


In order to measure the time constant , two
solutions of high and low temperatures were prepared.
The sudden step in temperature was used in comparison
to the actual measured values from each instrument. The
resultant graphs were fitted to the following equation for
transient temperature responses:
(1)
represents the temperature that the device will reach
after an infinite amount of time, while
represents the
initial temperature.
A beaker of water was brought to boiling
temperature on a hotplate, while a thermos was filled with
ice water. The RTD sensor was placed in the thermos

Copyright 2009 by ASME

until the reading stopped decreasing, at which point the


time was started and the probe was transferred to the
boiling water. The values displayed by the RTD were
recorded. This was repeated for a second set of data.
This process was then reversed, so that the probe was
initially placed in the boiling water and then transferred to
the ice water solution. This was repeated to obtain two
sets of values. The results for the RTD measurements
can be seen in figure1, and the values for the
thermocouple can be seen in figure 2.

Figure 3. Graph showing the linearization of RTD step up in temperature


exponential, trial 1

Figure 1. Plot showing measurements from RTD sensor for step up in


temperature

Figure 2. Measurements from RTD for step down in temperature

Copyright 2009 by ASME

Figure 4. Graph showing the linearization of RTD step up in temperature


exponential, trial 2

Figure 6. Graph showing the linearization of RTD step down in temperature


exponential, trial 2

Figure 5. Graph showing the linearization of RTD step down in temperature


exponential, trial 1

Figure 7. Thermocouple step up measurements in temperature

Copyright 2009 by ASME

Figure 8. Thermocouple measurements for step down temperature

Figure 10. Linearization of thermocouple step down in temperature

There were only three values that appeared


consecutively with enough room in between to analyze
and fit to the model of equation (1). However, since there
were less data points to fit, the linearization was much
more accurate, as illustrated in figures 9 and 10 with rsquared values close to one.

RESULTS
The average time response constant was
calculated to be 3.1295 1.6195s for the RTD rise, and
4.1665 3.56s for the fall. For the thermocouple, was
calculated to be 0.5620 0.1788s for the rise and 0.4228
2.1857s for the fall. The rise time to go from 10% to
90% of the step was calculated using the following
equation:
(2)
The rise time for the RTD was 6.925 1.955 seconds,
while the fall time was 9.1546 7.802 seconds. For the
thermocouple, the rise time was determined to me 1.235
0.393 seconds while the fall time was 0.929 4.802
seconds. All error bands are presented with a 95%
confidence interval.

Figure 9. Linearization of thermocouple step up in temperature

ANALYSIS
The equation (1) was linearized to make analysis
possible in MATLAB. The curves were plotted, and Excel
was used to create lines of best fit. By adjusting the
values manually, the r-squared value, or the coefficient of
linear correlation, changed to approach the number one.
The results of this linearization are illustrated in figures 3
and 4 for the rise in RTD, and figures 5 and 6 for the RTD
down. The closest
became the number used in the
actual Tau calculations. The instruments were not
provided with any bias error, and therefore the only error
resulted from precision bias, which was calculated using
the following equation:
(3)
The bias was calculated using a confidence interval of
95%. The error was relatively high, since only two
separate samples were used for the evaluation of the

Copyright 2009 by ASME

time constant. However, since there was relatively little


human interaction in acquiring data points, these errors
did not propagate into larger values during calculations.
Sources of error were associated with the placing of the
temperature probes, for both sensors. For the RTD, the
position of the probe in the ice water solution affected the
temperature measured. Since the ice floated to the
surface, placing the probe at varying depths would
produce inaccurate temperature readings. In the case of
the thermocouple, the temperature reading jumped over
100 Celsius at several points. This may have been due
to the contact end touching the sides or bottom of the
beaker, since liquid water is unable to reach a
temperature of over 100. Also, the color of the water was
slightly purple, and may have contained dissolved
chemicals from the temperature probe that caused the
boiling point to rise.
The thermocouple had a rise time of 1.235
0.1788s, which was approximately six times less than
that of the RTD. Thus, acquiring points that were
scattered between the temperature extremes was
difficult, as seen in figures 7 and 8.
CONCLUSION
The time constant of the resistance
temperature detector was evaluated to be 3.1295
1.6195s for the step up in temperature, and 4.1665
3.56s for the step down. For the thermocouple, the time
constant was calculated to be 0.5620 0.1788s for the
step up and 0.4228 2.1857s for the step down. The
time to rise and fall between 10% and 90% of the
temperatures was also determined. For the RTD, time of

rise was 6.925 1.955 seconds, while the fall time was
9.1546 7.802 seconds. For the thermocouple, the rise
time was determined to be 1.235 0.393 seconds while
the fall time was 0.929 4.802 seconds. These error
bands were all calculated for a 95% confidence interval.
The time constant did not seem to depend on the size of
the temperature step, but it did seem to be affected by
direction. The tau for stepping up was smaller than that of
stepping down.
The rise time for the thermocouple was indeed
much smaller than that of the RTD, which supports the
fact that each instrument was built for differing purposes.
The RTD is more useful in measuring temperatures in a
narrow range, making its response time less critical. The
thermocouple is more functional in measuring extreme
temperature ranges, approximately between -267C and
2316C [4].
REFERENCES
[1] Burns Engineering, Inc. RTD or Thermocouple?
http://www.burnsengineering.com/document/papers/rtd_vs_the
rmocouple.pdf
[2] Burns Engineering http://www.burnsengineering.com/faq/
[3]
Students
t-distribution
table
http://2.bp.blogspot.com/_0Lnn2oP30gU/TSkfPi_VSCI/AAAA
AAAAAKI/wkXYJGzMee8/s1600/tTableDistribution.jpg
[4] Nuclear Sensors and Process Instrumentation http://ultranspi.com/info_central/rtd.php

Copyright 2009 by ASME

You might also like