284  Measurement Automation Monitoring, Sep. 2016, no. 09, vol.

62, ISSN 2450-2855

Marek MAJDAK, Magdalena JAREMKIEWICZ

CRACOW UNIVERSITY OF TECHNOLOGY, INSTITUTE OF THERMAL POWER ENGINEERING
37 Jana Pawła II Ave., 31-864 Cracow

The analysis of thermocouple time constants

as a function of fluid velocity
Abstract thermocouple time constant is also very important for temperature
control systems and enables shortening the system response time.
In steady-state conditions when the fluid temperature is constant, there is In the processes that require maintaining precise temperature of
no damping and time lag so the temperature measurement can be
the fluid, the response time of the control system is crucial.
performed with a high accuracy. But when the fluid temperature is varying
rapidly as during the start-up, quite appreciable differences occur between
the exact and measured temperature because of the time required for the 2. Methods of determining the thermometer
transfer of heat to the thermocouple placed inside a thermometer pocket. time constant
The temperature of the fluid is one of the key parameters affecting the
proper operation of thermodynamic cycles, so the precise determination of
its value is very important. The speed of the response of control systems to Usually, the sheathed thermocouple is modeled as an element of
a temperature change is closely related to the time constant of the used the concentrated thermal capacity. It is assumed that the
thermocouples. The paper presents a significant impact of fluid velocity temperature of the thermometer is only a function of time, and
changes (in this case air) on the value of time constants of thermometers. temperature differences occurring in the thermometer are omitted.
For this purpose, the experimental study was carried out using sheathed Temperature changes of the thermometer in time T(t) are
thermometers with different diameters and hot junctions. The time described by an ordinary first order differential equation (first
constants determined for various thermometers are compared. order thermometer model)
Keywords: time constant, thermocouple, transient temperature
dT  t 
measurement, experimental analysis.   T t   Tf t  , (1)
dt
1. Introduction
where:
Temperature is one of the basic thermodynamic parameters of T – temperature indicated by the thermocouple, ºC,
the state. It is related to the average kinetic energy of vibration and Tf – temperature of the fluid, ºC,
movement of the molecules making up the system and is t – time, s,
a measure of that energy. It can be precisely specified only for the τ = mc/(αA) – time constant of the thermocouple, s.
state of thermodynamic equilibrium since in accordance with the
zeroth law of thermodynamics the temperature is the physical For thermometers with a heavy housing used for measuring the
quantity which determines the common property of the two temperature of the fluid under high pressure, the accuracy of the
systems that remain in the balance with each other. Knowledge of second order model is more adequate [15].
actual and accurate temperature is very important because it The solution of the differential equation (1) for the initial
significantly influences the course and character of processes condition:
occurring in nature. The fluid temperature is also one of the T  0  T0  0 (2)
parameters through which the flow of a substance can be
described, or efficiency of processes in power plants can be
is:
determined.
T  t   T0  t
Most books on temperature measurement concentrate on steady- u t    1  exp    , (3)
state measurements of fluid temperature [1-9]. Only a unit-step T f  T0  τ
response of thermometers is considered to estimate the dynamic
error of the temperature measurement. Little attention is paid to where u(t) – unit step response. The time constant of the
measurements of transient fluid temperature, despite the great thermocouple τ is a crucial parameter that characterizes this type
practical significance of the problem [10-12]. of a measuring device.
A sheathed thermocouple is widely used type of a temperature The simplest, but burdened with the greatest error, a way of
sensor. The thermocouple consists of two wires, which are made determining the time constant is a graphical method. It consists in
of different materials, joined at one end (junction end or leading the tangent to the curve illustrating the change in
measuring end) and forming a part of a system utilizing the temperature indicated by the thermocouple from the time that has
thermoelectric effect to measure temperature [13]. Thermocouples passed since a step change in temperature. The period of time
are divided into various types: R, S, B, J, T, E, K, N, due to the from the point of tangency to the intersection of the tangent with
different metals used in their construction. Thermocouple wires a straight drawn through a point corresponding to the temperature
must be isolated from each other along the entire length except the of the fluid is the time constant of the thermometer (Fig. 1).
measuring end. The thermocouple sheath protects the temperature A more accurate method of determining the time constant of the
sensor. The main advantages of thermocouples are small size, low thermometer is an experimental method using the method of least
heat capacity, wide measurement range, simplicity of construction squares. It was used to determine the time constant τ in Eq. (3).
and high reliability. However, like all real objects, thermocouples The values for the time constants are found by minimizing the
have drawbacks, which include instability of the measuring end function S
and the possibility of current flow beyond the thermocouple
circumference, which can lead to damage of the temperature N
S   um  ti   u  ti    min ,
2
sensor, and in extreme cases, to the destruction of computer (4)
hardware or problems with the correct operation of the system and i 1

its control.
The value of the thermocouple time constant allows determining where u(t) is the approximating function given by Eq. (3). The
when the temperature indicated by the measuring system is the symbol N denotes the number of measurements (ti, um(ti)). That is
actual temperature of the fluid [14]. Knowledge of the the sum of the squares of the deviations of the measured values
Measurement Automation Monitoring, Sep. 2016, no. 09, vol. 62, ISSN 2450-2855  285

um(ti) from the fitted values u(ti) is minimized. To eliminate This tunnel is utilized for calibration of anemometers and allows
partially random errors from the measurement data, a 9-point obtaining a known and steady velocity airflow over 100 mm × 100
digital filter can be applied [4]. Once the time constant τ has been mm test cross section. It uses the principle of operation of the
determined, it can be substituted into Eq. (4) to find the value for Venturi tube where the flow velocity of air is determined by the
Smin. difference between the pressure measured at the point of the inlet
The uncertainty in the calculated time constant τ is estimated to the tunnel and the point located in the measuring space (Figs 2
using the mean square error and 3).
The laboratory stand also includes a recorder that allows the
2
S min monitoring of air velocity, ambient temperature, barometric
SN  , (5) pressure and air humidity. All these parameters have the influence
N m on the pressure difference on the basis of which the air velocity in
the tunnel is determined.
where m is the number of time constants (m = 1 for Eq. (3)).
Based on the calculated mean square error SN, which is an
approximation of the standard deviation, the uncertainty in the
determined time constant can be calculated using TableCurve
software [16].

Fig. 3. Location of tubing connections

Fig. 1. Graphical method of determining the thermocouple time constant
During the experiments the tested thermometers were heated by
a fan heater and then cooled in the measuring space of the
3. Description of the laboratory stand laboratory stand at a known velocity of the air. The measurement
data was collected using the Ahlborn data acquisition system
construction (ALMEMO 2890-9).
To determine the relationship between the value of the time
constant of the sheathed thermocouple τ and the air velocity w, the 4. Analysis of the measurement results
open benchtop wind tunnel WT4401-D was used (Fig. 2).
As a result of the experiment time constants of the thermometers
(a) with different outer diameters were determined. Tests were carried
out on the sheathed thermocouples:
 Fe-CuNi (type J) with diameters: 1, 1.5, 2, 3, 4.5, 6 mm with
grounded hot junctions,
 Fe-CuNi (type J) with diameters: 1, 1.5, 2, 3, 4.5, 6 mm with
insulated hot junctions,
 NiCr-NiAl (type K) with diameters: 1.5, 3, 4.5, 6 mm with
uncovered hot junctions,
 NiCr-NiAl (type K) with diameters: 0.5, 1, 1.5, 3 mm with
grounded hot junctions.
For each thermometer, a series of temperature measurements for
(b) various velocities of air flowing through a wind tunnel was made.
Then, for each air velocity the time constant of the thermometer
was determined. The experimental data points collected for each
thermocouple were approximated by the least squares method to
the following formula [18]:

1
 ( w)  , (6)
ab w

where:
τ(w) – time constant, s,
Fig. 2. Wind tunnel used for determining the thermocouple time constant (a) overall w – velocity of air, m/s,
view (b) restrictive plate placement [17]: 1 − test chamber with opening for
the thermometer; 2 − differential pressure measurement; 3 − recorder,
a – constant, s-1,
4 − opening for the thermometer insertion, 5 − restrictive plate for adjusting b – constant, (ms)-1/2.
the velocity measurement range
286  Measurement Automation Monitoring, Sep. 2016, no. 09, vol. 62, ISSN 2450-2855

The best estimates for the constants a and b, with the 95%
confidence uncertainty in the results are presented in Tab. 1.

Tab. 1. Calculated values of searched coefficients a and b

The diameter of the sheathed Value of coefficients Value of coefficients

thermocouple a·10-2, s-1 b·10-2, (ms)-1/2
Fe-CuNi (type J) thermocouple with grounded hot junction
1.0 mm 1.83± 0.5 12.7± 0.7
1.5 mm 1.07 ± 0.1 6.80 ± 0.2
2.0 mm 0.467 ± 0.08 4.86 ± 0.1
3.0 mm 0.301 ± 0.06 2.71 ± 0.1
4.5 mm 0.234 ± 0.09 1.27 ± 0.08
6.0 mm 0.154 ± 0.08 0.863 ± 0.08
Fe-CuNi (type J) thermocouple with insulated hot junction
1.0 mm 3.49 ± 0.7 9.02 ± 0.8
1.5 mm 1.44 ± 0.4 6.42 ± 0.6
2.0 mm 1.08 ± 0.1 3.77 ± 0.2
3.0 mm 0.319 ± 0.1 2.54 ± 0.2
4.5 mm 0.14 ± 0.1 1.43 ± 0.1
6.0 mm 0.191 ± 0.06 0.850 ± 0.06
NiCr-NiAl (type K) thermocouple with uncovered hot junction
1.5 mm 0.696 ± 0.6 10.2 ± 0.9
3.0 mm 0.298 ± 0.1 2.98 ± 0.2
Fig. 5. Time constants τ of J-type sheathed thermocouple with grounded junction
4.5 mm 0.0857 ± 0.2 1.95 ± 0.2 end, J-type sheathed thermocouple with insulated junction end, K-type
6.0 mm 0.0365 ± 0.04 1.12 ± 0.2 sheathed thermocouple with uncovered junction end and K-type sheathed
NiCr-NiAl (type K) thermocouple with grounded hot junction thermocouple with grounded junction end with the outer diameter of 1.5 mm
as a function of air velocity w
0.5 mm 0.434 ± 0.06 2.22 ± 0.1
1.0 mm 2.1 ± 0.6 10.4 ± 1.0
The next diagram (Fig. 6) shows the difference between the time
1.5 mm 4.04 ± 0.3 5.69 ± 0.4
constants of J-type sheathed thermocouples with insulated
3.0 mm 12.8 ± 4.0 22.1 ± 5.0
junction ends with outer diameters: 1, 1.5, 2, 3, 4.5, 6 mm as
a function of the air velocity.
The time constant values obtained experimentally and the graph
of Eq. (6) for a J-type thermocouple with insulated junction end
with a diameter of 2 mm are shown in Fig. 4.

Fig. 6. Time constants τ of J-type sheathed thermocouples with insulated junction

ends with outer diameters of 1, 1.5, 2, 3, 4.5, 6 mm as a function of air
velocity w
Fig. 4. Time constants τ of a J-type sheathed thermocouple with insulated junction
end with the outer diameter of 2.0 mm as a function of air velocity w with The obtained results show that the fluid velocity significantly
95% confidence interval limits affects the time constant of the thermometer (Figs. 4-6). Also the
construction of thermometers has the impact on the value of the
The time constants of different types of thermocouples with an time constants (Fig. 5). The highest values of the time constants
outer diameter of 1.5 mm having different measuring ends are were obtained for the thermocouples with insulated hot junction,
shown in Fig. 5. slightly smaller for the thermocouples with grounded hot junction.
The considerably lower time constants values were obtained for
the thermometers with uncovered junction end.
Measurement Automation Monitoring, Sep. 2016, no. 09, vol. 62, ISSN 2450-2855  287

5. Conclusions Communications in Heat and Mass Transfer, vol. 28, 2001, pp. 1015-
1024.
The method presented in this paper for measuring the transient [12] Chau P.C.: Process control, A First Course with MATLAB.
temperature of a fluid can be used for the on-line monitoring of Cambridge University Press, Cambridge 2002.
[13] PN-EN 60584-1: Thermocouples − Part 1: Reference tables (IEC 584-
fluid temperature change in time. This method, where the
1:1995). December, 1997.
thermometer is modeled using an ordinary first-order differential
[14] Jaremkiewicz M., Sobota T., Taler D.: Measurement of transient fluid
equation, is appropriate for thermometers that have small time
temperature in installations of power plants. Pomiary Automatyka
constants. In such cases, the thermometer delay is small in
Kontrola, vol. 55, no. 05, pp. 288-291, 2009 (in Polish).
comparison to the changes of the fluid temperature. When the
[15] Jaremkiewicz M., Taler D., Sobota T.: Measuring Transient
thermometer delay is big, it is more appropriate to consider the
Temperature of the Medium in Power Engineering Machines and
thermometer as a second-order inertia device. Substantial stability
Installations. Applied Thermal Engineering, vol. 29, 2009, pp. 3374-
and accuracy of the computed actual fluid temperature from the 3379.
measured thermometer temperature can be achieved by using a 9- [16] TableCurve 2D v.5.0, Automated Curve Fitting&Equation Discovery.
point digital filter. The technique proposed in this paper can also AISN Software Inc., 2000.
be used when the thermometer time constant is a function of fluid [17] WT4401-S & WT4401-D Benchtop Wind Tunnels. Omega, Stamford,
velocity. CT, USA, www.omega.com.
[18] Jaremkiewicz M.: Inverse heat transfer problem encountered in
6. References measurement of transient fluid temperature. Publishing of Cracow
University of Technology, Cracow 2011 (in Polish).
[1] Nicholas J.V., White D.R.: Traceable Temperatures. An Introduction _____________________________________________________
to Temperature Measurement and Calibration. Second Edition. Wiley, Received: 04.06.2016 Paper reviewed Accepted: 01.08.2016
New York2001.
[2] Michalski L., Eckersdorf K., McGhee J.: Temperature Measurement.
Wiley, Chichester 1991.
[3] Wiśniewski S.: Temperature Measurement in Engines and Thermal Marek MAJDAK, MSc, eng.
Facilities. WNT, Warszawa 1983 (in Polish).
He graduated from full-time studies in Energy
[4] Taler J.: Theory and Practice of Identification of Heat Transfer Technology with specialization in Renewable Energy at
Processes. Zakład Narodowy imienia Ossolińskich, Wrocław 1995 (in the Faculty of Mechanical Engineering of Cracow
Polish). University of Technology. He is working at the Faculty
[5] Kabza Z., Kostyrko K., Zator S., Łobzowski A., Szkolnikowski W.: of Mechanical Engineering at Cracow University of
Technology as a research assistant.
Room Climate Control. Agenda Wydawnicza, Pomiary Automatyka
Kontrola, Warszawa 2005 (in Polish).
[6] Littler D.J., Kirkby F., Johnson H.E., Myerscough P.B., Davies E.J.,
Wright W.: Modern Power Station Practice, Vol. F: Control and
Instrumentation. Third Edition. Elsevier, Amsterdam 2008. e-mail: marek.majdak@mech.pk.edu.pl
[7] Childs P.R.N.: Practical Temperature Measurement. Buterworth-
Heinemann, Oxford 2001.
Magdalena JAREMKIEWICZ, PhD
[8] Gerashchenko O.A., Gordov A.N., Lakh V.I., Stadnyk B.I., Yaryshev
N.A.: Temperaturnye Izmereniya. Naukova dumka, Kiev 1984 (in She graduated from the Cracow University of
Russian). Technology in 2006. She deals with issues of inverse
[9] Han J.C., Dutta S., Ekkad S.V.:Gas Turbine Heat Transfer and heat transfer. She has published 30 papers in scientific
journals and several chapters in monographs. She is also
Cooling Technology. Chapter 6: Experimental Methods. the author of the monograph Inverse heat transfer
Taylor&Francis, New York, London 2000. problem encountered in measurement of transient fluid
[10] Székely V., Ress S., Poppe A., Török S., Magyari D., Benedek Zs., temperature (2012).
Torki K., Courtois B., Rencz M.: New approaches in the transient
thermal measurements. Microelectronics Journal, vol. 31, 2000, pp.
727-733.
[11] Crocker D.S., Parang M.: Unsteady temperature measurement in an e-mail: mjaremkiewicz@pk.edu.pl
enclosed thermoconvectively heated air. International