IRJET - Control and Analyze of TCPTF

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 03 | Mar 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 1894
CONTROL AND ANALYZE OF TCPTF
Mr. P. NARENDRA ILAYA PALLAVAN*, S.N. MADHUMIDHA$, V. PAVITHRA$
1Assistant Professor, Dept. of Electronics and Instrumentation Engineering, Bannari Amman Institute of
Technology, Tamilnadu, India
2,3Student & Bannari Amman Institute of Technology, Tamilnadu, India
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract - The method of study and experimental researches
of the error of method of the thermocouple with controlled
profile of temperature field along the main thermocouple are
considered in this project. Thermocouple error is potential for
thermocouple to lose accuracy over time, moving in positive
direction. Thermocouple error is more common when
thermocouple has been operating at higher temperaturefora
prolonged period. Thermocouple operating within normal
operating parameter or atambienttemperature maystill drift
much slower. While with this project a thermocouple is used
and its positive and negative terminals are connected to the
DAQ and they are interfaced with PC for collectingthereading
obtained from the thermocouples for the error analysis.
Thermocouples are kept inside the control profile for the
analysis of drift and the thermostat is kept at a set point of
800ºC. The responses are taken for the thermocouples. Then
the degradation is done for the thermocouple and again the
drift is analysed and this process isrepeatedforcertainperiod.
Experimentally determined values of this method are
compared to the theoretical estimations done using
Thermocouple with Controlled Profile of Temperature Field.
Key Words: DAQ, Thermocouple, Error analysis.
1. INTRODUCTION
Type K (chromel- alumel) is the most
common general-purpose thermocouple. When a
thermocouple is used to measure the live temperatureof an
environment, it's expected that the voltage
acquired doesn't change if the temperature of the
environment is constant. If the voltage changes with
time when the temperature oftheenvironmentisconstant it
could be a source of error in thermocouple measurement
called drift. Due to drift, the thermocouple loses accuracy
over time.
Temperature is one of the most commonly
measured physical quantities. Statistical data show that
about of all measurements in industry, the percentage of
temperature measurements is 40 %. In some branches of
industry this percentage may be even considerably higher,
particularly in power industry it equals about 70 %.
However, for existing measuring methods technological
progress is challenging and expands the need for
measurements in new conditions. At the same time,
requirements for accuracy of measurements are increasing
in areas and industry, where temperature has been
measured for a long time.
A lot of effort has been directed to correct the error
due to drift of conversion characteristics of thermocouple
using periodic verification or calibration of thermocouple
and drift prediction. However, this method does not lead to
considerableimprovementofaccuracyduetothermoelectric
inhomogeneity of thermocouple electrodes.
2. LITERATURE SURVEY
[1] Davor Zvizdic, Tomislav Veliki, et. al., proposed that
techniques for inhomogeneitytesting .Inthisvariationof the
thermovoltage recordedduringmeasurement is employed in
calculation of the uncertainty of the calibration.
[2] Holmsten, M., Ivarsson, J., Falk, R., Lidbeck, M., Josefson,
L.-E, et. al., proposed a method that uses the thermoelectric
inhomogeneity of wires to measure uncertainty when using
thermocouples.Toassessthermoelectricinhomogeneityover
greater lengths, it's necessary to adopt a unique technique.
Therefore, an apparatus with a brief, movable heating zone
has been founded and evaluated.
[3] Hill, K.D., Gee, D.J., et. al., analysed the
inhomogeneity withinthe seebeck coefficient isa functionof
position along a thermocouple wire frequently dominates
the uncertainty budgets of thermocouple calibration and
use.It explore how the inhomogeneity impacts the
calibration uncertainty attainable with the varied thermal
sources used for the calibration of thermocouples.
[4] Su Jun, Kochan, O, et. al., examines the explanations of
error becauseof thermoelectricinhomogeneityof electrodes
of thermocouples acquired during prolonged use and
therefore the neural network method of error correction
supported a generalization of verification leads to several
temperature fields and also the tactic of investigating the
impact of fixing the speed of the conversion characteristic
drift of thermocouple on error correction and results of this
investigation.
[5] Glowacz,A., Glowacz, A., Korohoda, P., et. al., predicated
the study of thermal images of the motor. This
method is useful fordefense of electricmotor.Moreover, this
method is used to diagnose equipments in steelworks and
other industrial plants.
[6] Webster, E.S., White, D.R., Edgar, H, et. al., describes a
linear-gradient furnace and a thermocouple homogeneity
scanner that, together, measure changes within the Seeback
coefficient as a function of your time and temperature. The

furnace first exposes the test thermocouple to any or
all temperatures within the range spanned by the furnace
gradient.
[7] Koci, V., Koci, J., Korecky, T., Madera, J., Cerny. R, et. al.,
This employes a combination of experimental measurement
and computational modeling. Within the experimental part,
cement mortar specimen is heated during a laboratory
furnace to 600°C and therefore the temperature field inside
is recorded using built-in K-type thermocouples connected
to an information logger.
[8] Glowacz, A., Glowacz, A., Glowacz. Z, et. al., this article
present a diagnostic method of incipientfaultdetection.This
approach relies on pattern recognition. Itusesmonochrome
thermal images of the rotor with the applying of a
region perimeter vector and a Bayes classifier. The
investigations are administered for DC motorwithoutfaults.
and motor with shorted rotor coils
[9] Habisreuther, T., Elsmann, T., Pan, Z., Graf, A., Willsch, R.,
Schmidt, M.A,et. al., This method reports on a brand
new quite temperature sensor operating over a very large
temperature range at high monitoring speeds. The gratings
operate up to 1900 °C, which is that the highesttemperature
determined using Bragg grating to this point, and
permit signal processing with a temperature resolution
better than ±2 K. The sensor uniquely provides fast dynamic
temperature monitoring at an unprecedented rate of 20 Hz.
[10] Kochan, О., Kochan, R., Bojko, O., Chyrka, M, et. al., A
Thermocouple with controlled temperature field was
proposed to address this error. An information-measuring
systemtoperformpropermeasurements,measurement data
acquisition and collection to construct mathematical models
is proposed. They showed that the coefficient of penetration
of the temperature field of the measured object is about
0.04.
3. PROPOSED WORK
3.1 EXPERIMENTAL PROCEDURE
The three set of thermocouples whose positive and
negative terminals areconnectedtoDAQandinterfacedwith
PC, the readings are collected from the thermocouplefor the
analysis of the error. The error analysis is done with the
thermocouples which are kept inside the controlled profile
at a set point of 800ºc. The n the set point is reducedto600ºc
with this the down time and the settling time is noted.
The set point is increased to 800ºc for the
degradation process. Since the error of the thermocouplesis
much larger than the error of method of TCPTF, the direct
measurement of this error is impossible. That is why it is
proposed to use the relative method of measurements to
solve the problem. It is based on measuring the changes in
temperature which occur when changing the power of the
heater.
Fig -1: Experimental setup of TCPTF
Thus thesystematiccomponentofthemeasurement
error will be mutually adjusted. Since the outcome of the
experiment may be affected by some random temperature
changes of the environment during the experiment time, the
measuring junction of temperature sensor is placed into the
passive thermostat. Its temperature is measured by an
additional resistance temperature detector. During the first
stage of the experiment, the heater is off. It is necessary to
measure the temperatures of sensor several times and to
check if their temperaturesdonotchange morethanrandom
error of measurement. During the second stage of the
experiment the heater is on. This causes the temperature
change measured by temperature sensor. After a while the
heat flux from the heater will changethetemperaturesofthe
measuring junction of temperature sensor and its passive
thermostat.
3.2 DATA ANALYSIS
3.2.1 SMOOTHING
Smoothing is a sign processing technique generally
used to remove noise from signals. In smoothing, individual
factors which might be higher than the right away adjoining
points are reduced and factors which are decrease than the
adjacent factors are elevated in order that the data factorsof
a sign are modified. Discovering important patterns inside
the statistics whilst leaving out noise, outliers and different
inappropriate statistics. It also removes unwanted spikes,
trends and outliers from a signal.
Filtering is used to perform smoothing. The purpose
of smoothing is to generate slow changes in value so that it's
easier to see trends in the data. Smoothingoperationscanbe
implemented more than as soon as, In a few instances it is
able to be beneficial if there is a brilliant deal of high-
frequency noise within the signal.

When the smooth width is larger, the noise reduction
is greater. The easy width depends upon the width and
shape of the sign and the digitization interval. For peak-kind
alerts, clean ratio is the important aspect.Smooth ratioisthe
ratio between the width of the smooth signal and the wide
variety of points in the half of-width of the height. In trendy,
growing the smoothing ratio improves the signal-to-noise
ratio however reasons a reduction in amplitudeandboomin
the bandwidth of the peak.
3.2.2 FFT
If the signal has high-frequency noise, FFT
filtering method is used. This filter removes all the high
frequency noise, leaving the true signal.
When the frequency of noise is higher than the true
signal FFT low-pass filter is used.Byusingparabolic window
it removes the high- frequency components. Frequencies
more than the cutoff frequency will be discarded
3.3 OUTLIER DEDUCTION
3.3.1 DESCRIPTIVE STATISTICS
Descriptive facts are regularlygeneratedasa primary
step in performing a statistical evaluation.Theyaremeantto
provide a concise examine a group of facts factors, thru such
information as imply and variance.
Descriptive analysis is one of the statistical
information evaluation strategies which constantly being
completed prior to undertaking any statistical checks or
more complicated modeling. It is applied to summarize the
records by describing and characterizing the records.
Descriptive analysis is regularly used to measure statistics
statistical distribution and relevant tendency. A statistical
distribution specifies the amount of occurrences of the
chosen records supported unique categorization. The
quantity of occurrences can also be exact employing a
percent fee for every category. A statistical distribution is
illustrated via desk or graphical visualizationlikelinecharts,
pie charts and bar charts. Meanwhile central tendency
describes the middle values of the chosen statistics which
typically represented the use of mode, suggest and median
values.
3.3.2 GRUBBS TEST
An outlier is statistically an commentary which is
numerically remoted from the relaxation of the information.
To examine whether there's an outlier in a records set from
repeated measurements, Grubbs test is employed.
Sometimesalertsarecontaminated withverytall,slim
“spikes” or “outliers” occurring random periods and with
random amplitudes. This type of interference is hard to do
away with the use of the above smoothing methods without
distorting the signal. Many filters are sensitive to outliers.
A filter which is closely associated with the mean
filter is that the FFT filter. However, a FFT filter out, which
replaces every factor in the signal with the mean adjacent
points can completely dispose of narrow spikes, with little
trade inside the signal, if the width of the spikes is slightly
one or a few factors and as much as or less than mean.
3.3.3 HYPOTHESIS TESTS
Hypothesis tests arefrequentlyusedtodegreethesame
old of sample parameters or to check whether or not
estimations for two samples on a given parameter are equal.
Hypothesis testing with unknown parameters in null
hypothesis and alternative hypothesis has been widely
applied in many fields such as the signal processing system,
the financial services and the wireless telecommunications
system.
When the prior distributions of the unknown
parameters are known completely, this method is appliedin
order to get the optimal estimation and detection. Other
hypothesis testing methods solve the problem without
considering the fact that the prior distributions are partly
known. However, in practical, there are always some errors
in prior statistical information, which are caused by small
sample size, noises, model with uncertainties, etc.
3.3.4 T-TEST
The one-sample t-Test determines the average of a
sample taken from a normally distributed population is
steady with the hypothetical value for a given confidence
level.
3.4 BLOCK DIAGRAM OF THE PROPOSED SYSTEM
Fig -2: Block diagram of proposed system
The block diagram of TCPTF is in the top part of Fig.2. It
shows all the sensors in the experiment - thermocouples(the
MTC and the first zone of heating TC1) and the RTDs.During
the first stage of the experiment, all the heaters are off. The
temperature of all the sensors is the same as in the ambient
air. It is necessary to measurethe temperaturesofallsensors

several times and to check if their temperatures do not
change more than random error of measurement channels.
During the second stage of the experiment the heater
of zone 1 is run at full power. This causes the temperature
change of the first heating zone, measured by TC1. After a
while the heat flux from the heater of zone 1 will change the
temperatures of the measuring junction of MTC and its
passive thermostat. One must periodically measure these
temperatures until they are steady (transition process of
heating comes to the end). Since the transition processes are
long-lasting during the temperaturechange,itisnecessaryto
take into account possible changes of temperature in the
thermostat of their reference junctions during the
temperature measurement with thermocouples.
4. RESULT AND DISCUSSION
For RTD measurement there is used the
potentiometer circuit with its four-lead arrangement. The
results of the measurements and calculations is described
below.
The result obtained from smoothing the data is shown in
figure 3.
Fig -3 : Smoothing
Thus, the noise, unwanted spikes and trends is
removed from the collected data through smoothing.
Fig -4: Grubbs test
From the figure 4,mean and standard deviation for
smoothened data is calculated through Grubbs test.
Fig -5: outlier deduction
From the figure 5,the outlier for the data is suspected
Figure 6: Histogram
The number of samples analysed using TCPTF method is
shown in figure 6.

5. CONCLUSION
The proposed technique is primarily based on
measuring the changes in temperature during the relative
measurements. It reduces the error because of the heat flux
from the auxiliary heaters of TCPTF. Thus the error values
found the use of TCPTF technique is three times less than the
error of thermocouples that may be used for calibration of
TCPTF. So, the proposed method of the analysis may be used
for the sensible purposes. Error also can be decreased by
means of growing the thermowell heat transfer location by
way of making use of both grooves or thread, or mount extra
fins at the give up of the thermowell.Thisapproachestimates
the impact of thermoelectric inhomogeneity on the
uncertainty of the calibration procedure. It is a great
calibration practice to test thermocouples for homogeneity
for the duration of calibration process
REFERENCES
[1] Davor Zvizdic, Tomislav Veliki., “Testing of
thermocouples for inhomogeneity”, 2006.
[2] Holmsten, M., Ivarsson, J., Falk, R., Lidbeck, M., Josefson,
L.-E “Inhomogeneity measurements of long
thermocouples using a short movable heating zone”,
2008.
[3] Hill, K.D., Gee, D.J., “Quantifying the calibration
uncertainty attributable to thermocouple
inhomogeneity”, 2013.
[4] Su Jun, Kochan, O., “Investigations of thermocoupledrift
irregularity impact on error of their inhomogeneity
correction”, 2014.
[5] Glowacz, A., Glowacz, A., Korohoda, P., “Recognition of
monochrome thermal images of synchronous motor
with the application of binarization and nearest mean
classifier ” , 2014.
[6] Webster, E.S., White, D.R., Edgar, H., “Measurement of
inhomogeneitiesinMIMSthermocouplesusinga linear-
gradient furnace and dual heat-pipe scanner”, 2014.
[7] Koci, V., Koci, J., Korecky, T., Madera, J., Cerny, R.,
“Determination of radiative heat transfer coefficient at
high temperatures using a combined,experimental-
computational technique ” , 2015.
[8] Glowacz, A., Glowacz, A., Glowacz, Z., “Recognition of
thermal images of direct current motor with
application of area perimeter vector and bayes
classifier”, 2015.
[9] Habisreuther, T., Elsmann, T., Pan, Z., Graf, A., Willsch, R.,
Schmidt, M.A., “Sapphire fiber Bragg gratings for high
temperature and dynamic temperature diagnostics ” ,
2015.
[10] Kochan, О., Kochan, R., Bojko, O., Chyrka, M.,
“Temperature measurement system based on
thermocouple with controlled temperature field ” ,
2019.

IRJET - Control and Analyze of TCPTF

Related slideshows

More Related Content

IRJET - Control and Analyze of TCPTF