Methods of measuring flame temperature

A seminar report on

Methods of measuring
flame temperature

Submitted by
Shainath R. Kalamkar
M Tech 1st year Roll No. 09TH04F

To
Dr. P. Mohanan
Professor, Mechanical Engineering Department
NITK,Surathkal

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 1

Methods of measuring flame temperature

ABSTRACT
Adiabatic flame temperature which we are calculating theoretically does not reflect the actual
temperature of the flame. So there is need to measure the accurate temperature of the
flame. There are some methods of measurement of temperature of the flame this seminar
contains the methods used in measurement of temperature of the flame. It includes
working principle of the method, techniques involved, instrumentation, calculations,
limitations and future challenges in front of the flame temperature measurement system.

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 2

Methods of measuring flame temperature

INDEX
Abstract.2

1. Introduction4

2. Methods...5

2.1

Using thermocouples...5

2.2

Temperature measurement using a fine wire compensated

thermometry.7

2.3

Mach-Zehnder Interferometry10

3. Conclusion17

4. References.18

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 3

Methods of measuring flame temperature

CHAPTER 1

INTRODUCTION
Study of the temperature field of a flame is crucial to the design of a burner, since
the local temperatures affect the species reaction rate, equilibrium concentration and even the
emission characteristics. Besides, data of the temperature field of the premixed flame is the
key in determining the heat transfer rate when the heat released from combustion is
transferred to a receiving medium such as an impingement plate. Thus, it is one of the big
concerns in both experimental and as a usual practice, local flame temperatures are measured
by inserting thermocouples into different locations of the flame and the obtained readings are
then used to construct the entire temperature field. However, this method has the drawbacks
that temperatures at different locations are not obtained simultaneously (in other words, not
under the same operating conditions), and the flow field is disturbed by the inserted
thermocouple. In addition, the obtained readings need to be compensated against the effects
brought by radiation and Thermo-Chemical-Heat-Release (TCHR). These drawbacks lead to
the search of reliable non-contact methods of flame temperature measurement.

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 4

Methods of measuring flame temperature

CHAPTER 2

METHODS
Following methods of flame temperature measurement we will discuss are

1 Using thermocouples.
2 Temperature measurement using a fine wire compensated
thermometry.
3 MachZehnder Interferometry.

1 Using thermocouples
Thermocouples have been used extensively throughout industry to make
temperature measurements in a variety of engineering situations.

Fig.1.1 Flame probing through the years, &showing the increase in complexity
of probing techniques.[3]

Such devices have been relied on extensively to measure the temperature of hot gases
present inside laboratory and large scale outdoor flames. They are the most widely used
thermal probes for flames. Thermocouples are the temperature measuring devices of
choice, because they have relatively fast response times if appropriately sized, can

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 5

Methods of measuring flame temperature

withstand high temperatures by appropriate material selection, are rugged, and are low
cost.

Probes
Thermocouples can be either bare junction, as shown in Figure1.2, or with the junction
enclosed by a metal sheath which provides electrical isolation from the environment, as
shown in Figure 1.2. In the latter case, thermal energy must conduct through the metal
sheath and insulation before a change in thermoelectric voltage at the circuit can develop.
Figure (a) and Figure (b) are not to scale, and in fact, the size of the bead for the bare
junction is larger than the diameter of the sheath for the enclosed configuration in this
study. However, in compensation for the larger thermal mass, there is no thermal
insulation associated with the bare junction to insulate the junction from the environment,
as is the case with the sheathed thermocouple.

Fig.1.2 Probes

The basic setup used during the acquisition of thermocouple data involved attaching
thermocouple probes to an analog/ digital low voltage data acquisition board. The board was
connected to a high resolution data acquisition card (PCI 6025-E) with built-in analogue-todigital converter. The signal was acquired, saved, and analyzed using special software. The
setup had the capability of visualizing the signal in real time and performing all the postMTECH THERMAL ENGINEERING NITK SURATHKAL

Page 6

Methods of measuring flame temperature

processing at the same time the signal was being acquired. The data collection system was
equipped with an internal reference to avoid the need to set up and maintain an ice bath
reference junction. [2]

2. Temperature measurement using a fine wire

compensated thermometry
The techniques for measuring the flames temperature fall into two categories: intrusive and
non-intrusive technique. In our study, the measurement technique by B type thermocouple FWCT (Fine Wire Compensated Thermocouples) has been used. To overcome the drawback
involved in the use of probes compensated thermometry method is used. So that the result
obtained by this method are satisfactory. [4]

Experimental setup
In this study thermocouple type B was used; figure2.1 shows this type of thermocouple. The
weld heat of these thermocouples is realized between two wires of small-diameter (10 500
m) to increase the temporal and spatial resolution of the measurement. The wires are fixed
to pins more resistant to flow. Each pin is made of the same alloy as the corresponding part of
the wire. When the temperature of the environment is high, the radiant loss becomes
important, thus the temperature measured by the hot junction is less than the environment
temperature. These losses have been estimated and discussed by many authors, (Paranthon
& Lecordier, [1996], Larras [2000], Ikegami & al. [2003], Moss & al., [2007]. Their
estimates by one model, can correct the difference between thermocouple temperature and
real gas temperature. In addition to that, the temporal resolution of the signal thermocouple is
a very important parameter for turbulent flow study. The use of this technique requires an
accurate knowledge of heat transfer between the thermocouple and the flow, the physical
characteristics of the probe materials, and its geometrical dimensions. That compensation is
needed for two reasons: The first relates to the temperature displayed by a thermocouple,
which is not necessarily the environment gas temperature, as a result of radiation losses from
high temperature thermocouples. For the gas temperature, it is necessary to estimate and
MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 7

Methods of measuring flame temperature

correct those losses. The second reason relates to the time response of the thermocouple,
which is limited by the inertia of two wires. The digital compensation is then used.

Fig.2.1 Thermocouple

Knowing the time response characteristic of the thermocouple, i.e. time constants , we can
reconstruct the original temperature signal. The method, which takes into account the
convective and radiative effects, is directly extracted from the energy balance. It is
determined by introducing the gas temperature and the convective time constant.

The equation can be obtained simultaneously offsetting radiation losses and inertia. This
equation is mainly a function of temperature measured by the wire and the Nusselt number.
The compensation of inertia thermocouple requires an iterative process. The B type
thermocouple (Pt30% Rh-Rh Pt6%) is used in the temperature range from 700 to 2000 K,
where it owns good signal amplitude (about 9 v / K). A diameter of 50 m was chosen to
MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 8

Methods of measuring flame temperature

be a good compromise between a low-inertia thermocouple and sufficient mechanical
strength. The theoretical values for uncoated thermocouples were calculated from equations
proposed in the literature. [4]

Results and conclusion

In unburned gas, fluctuations have low values. On average, they are in the range of 20 C, i.e.
1.3% of the temperature of the hot gas. In areas of interaction of unburned gas with the
combustion zone, one can notice that fluctuations is the highest, they are between 285 and
410 C (Figure 2). The results obtained using compensation was very consistent. In this
figure, it is clear that the compensation of a digital signal temperature tends to accentuate the
gap 1000C. It would therefore be interesting to interpret the spatial structure characteristic
fluctuations depending on their operating conditions.

Fig.2.1 Profile of temperature and RMS (C: Compensed and N.C not compensed).

In figure 2.2, a validation of measurements made by the technique of FWCT was carried out
by making a comparison with the methods of optical measurements (Rayleigh scattering) was
carried out. The results of Rayleigh scattering are those obtained by Boukhalfa [1988]. This
comparison shows a good agreement.
MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 9

Methods of measuring flame temperature

Fig.2.2 Comparison of temperature measurement with FWCT and Rayleigh

Scattering

An experimental procedure with post processing was developed to determine the thermal
field in premixed turbulent flames. This procedure was carried out using the Fine Wire
Compensed Thermometry (FWCT) technique. The method, which takes into account the
convective and radiative effects, is directly extracted from the energy balance. The addition
of catalytic effect in the energy balance takes good results. The comparison between FWCT
and Rayleigh scattering results, shows a good agreement. [4]

3. MachZehnder Interferometry
The temperature field of a premixed slot laminar flame jet is visualized by studying the
interferogram of the flame using MachZehnder Interferometry. The introduction of a slot
burner with a large aspect ratio enables the three-dimensional temperature field of the flame
to be observed directly from the fringe patterns of the obtained two-dimensional
interferogram. Subsequently, the temperature field of an open flame jet at a specified
Reynolds number or equivalence ratio is studied by varying one parameter while keeping the
other unchanged. At the Reynolds number of 400, temperature fields of the flame jet at
different equivalence ratios ranging from 1.0 to 2.0 are studied and compared. Similarly, at an
equivalence ratio of 2.0, temperature fields of the flame jet at different Reynolds numbers
ranging from 400 to 900 are investigated. In the present study, a non-contact method is
MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 10

Methods of measuring flame temperature

successfully developed to measure the temperature field of a premixed butane/air slot laminar
open flame jet. [1]

Experimental setup
A Reference Beam Interferometer (RBI) with the setup of a MachZehnder
Interferometry as shown in Fig1 is used to study the temperature field of the threedimensional premixed slot flame jet in the present study. A collimated laser light is split by
the beam splitter into two different light beams: one traveled along the original path and
another traveled vertically in the cross-flow direction. The path of each beam is then
converted vertically by the mirror, which is placed normal to the optical path. Consequently,
the two beams intersected at another beam splitter. The interferometry pattern is formed on a
diffuser (ground glass) and then imaged by a convex lens placed after the beam splitter and
the diffuser, onto the CCD sensor of a digital camera. Ideally, there should be no background
fringes observed from the diffuser when the light beams are parallel and the refractive index
of the media remains constant in both light paths of the two laser beams. Once the high
temperature premixed slot flame jet is inserted into the path of one of the two light beams,
highly contrast fringe patterns are projected into the digital camera. Digital photography at 30
frames per second makes it possible to observe real time temperature field. Consequently, the
temperature field of the flame is obtained via the GladstoneDale equation. The Mach
Zehnder Interferometer employed in the present study and shown in Fig.1 consists of a 5 mW
HeNe laser source with a wavelength of 632.8 nm, a collimator, two 50.8 mm cube beam
splitters, two flat mirrors and one convex lens. The distance from the beam splitter to the
reflecting mirror is maintained large enough to place the slot flame. The premixed slot flame
jet is emitted from a rectangular nozzle with a large aspect ratio of 32.3 mm _ 1.9 mm as
shown in Fig.1.

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 11

Methods of measuring flame temperature

Fig3.1 Setup of the MachZehnder Interferometry for temperature field measurement of a premixed slot
flame jet.

The butane-gas and compressed-air are metered and then premixed in a brass cylinder before
being conveyed via a 90 mm long stainless steel tube to cylindrical brass equalizing chamber.
The equalizing chamber is fully filled with very small stainless steel beads to ensure a
uniform flow and avoid flashing back of the flame. The premixed butane/air mixture is then
supplied to the rectangular flame holder, where it is ignited to produce slot flame jet. The
Reynolds number and equivalence ratio are determined from the two flow rates. In the
present study, the Reynolds number of the flame is adjusted from400 to 900. Due to
environmental concerns, it is common that only lean mixtures are burnt in real burners.

Fig 3.2 Rectangular burner for production of premixed slot ame.

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 12

Methods of measuring flame temperature

However, in the present study, the equivalence ratio is varied from 1.0to 2.0 which seem to be
rich. The rationale is that the mixture will continue to mix with ambient air after exiting from
the nozzle resulting in an increasingly leaner mixture. It is found that if the equivalence ratio
is low, the flame is unstable. The equivalence ratio of 1.02.0 allows the experiment to be
conducted for the 400900 range of Reynolds number of the flame. The rectangular nozzle is
inserted into one path of the two parallel light beams with the shorter edge normal to the light
beam, and each picture of fringe patterns is recorded under specified Reynolds number and
equivalence ratio. The temperature field calculated with the aid of the recorded fringe
patterns is compared with that obtained via thermocouple measurement at the same Reynolds
number and equivalence ratio. In every experiment, the temperature, relative humidity and
pressure of the environment are measured because the refractive index of the surrounding air
is used as the reference value and is important in the calculations of the refractive index of
the flame .Similar tests are conducted with a square nozzle of7.5 mm _ 7.5 mm at a Reynolds
number of 800. A high equivalence ratio of 2.0 is used to allow the flame to operate stably at
this Reynolds number. By comparing with the results obtained with the slot flame jet of large
aspect ratio, it is shown that, the temperature field of a three-dimensional flame can be
obtained with good accuracy by the MachZehnder Interferometry technique. [1]

Temperature measurement
When the light wave generated from a coherent light source passes through a hot flame, the
phase information of the light wave is varied according to variation of the refractive index of
the flame. The extent of this change, compared with the light beam in the reference state, is
determined by the refractive index field and its length. Therefore, the relative phase
difference, (x, y), of the two interfering light waves as shown in Fig.1 can be described as:

In Eq. (1), n is a constant that denotes the ambient refractive index at the time of the
fringe pattern is captured and n(x,y, z) denotes the local refractive index along the light
path through the flame, while L is the distance traveled by the light beam when it passes
through the flame. For a slot flame produced from a rectangular nozzle with a large aspect
ratio, the end effects at both ends of the longer edge are negligible. The refractive index is

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 13

Methods of measuring flame temperature

assumed to be constant along the light beam path (i.e. z-axis), and the condition can be
simplified to become a two-dimensional flame problem. Thus, Eq. (1) can be simplified as

Phase information in the fringe map is denoted by the fringe number,

of each fringe. Thus, the refractive index of a certain point in the xy plane can be
described as

As soon as the local refractive index is determined from the fringe pattern, the corresponding
temperature can be deduced by applying the GladstoneDale equation:

In Eq. (4), T(x,y) is the temperature of a certain point of coordinates (x,y) in the xy plane,
and T is the ambient temperature not affected by the flame.[1]

Error analysis
Temperature eld measurement of a premixed ame is based on the assumption that the
refractive index of an air/fuel mixture during combustion is similar to that of air. Thus, the
difference in refractive index of air and ame/combustion-gases is one of the error sources.
Qinet al. suggested that the average error caused by the chemical composition of a premixed
gas-red circular open ame was around 2.3% and 2.0%, at the equivalence ratio of 2.0 (i.e. a
rather fuel-rich ame) and 1.0 (i.e. stoichiometric ame), respectively. In the present study,
the applied equivalence ratios are within the range of 1.02.0; so that the error brought by
gas/ame composition should be limited to around 3%.The other major source of error in the
MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 14

Methods of measuring flame temperature

present study is due to the end effect of the rectangular burner. It is assumed that no
temperature variation occurred in the ame along the light path of the laser beam (i.e. z-axis),
therefore the refractive index as well as the temperature eld in the xy plane can be
calculated directly from the fringe map. However, the fringe patterns on the fringe map
includes both the temperature eld information along the laser beam path and that emitted
from the end of the slot burner, where the ame temperatures are signicantly lower due to
the rapid heat loss to the ambient air. Error caused by such end effect should be lower if the
lower temperature region is smaller, which suggests the use of a rectangular burner with a
large aspect ratio with its longer edge parallel to the z-axis. [1]

Result and discussion

Temperature elds as shown in Fig.3.1 are obtained from two diferent burners, one is a
square burner of 7.5 mm x7.5 mm and the other is a rectangular burner of 32.3 mm x1.9 mm.
Both burners are operating at a Reynolds number of 800 and an equivalence ratio of 2.0,
temperatures along a certain horizontal line where thermocouple measurement is carried out
are compared with those on the same horizontal line calculated from the fringe patterns. The
measured ame temperatures are rst corrected for convection and radiation efects
according to the method suggested by Bradley and Matthews [28].It is found that a better
agreement is obtained with the rectangular burner (i.e. Fig. 2(d)) than with the square burner
(i.e. Fig. 2(c)). Thus, it is suggested that the error caused by the end efect of the rectangular
burner should be decreased with its increasing aspect ratio.

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 15

Methods of measuring flame temperature

Fig. 3.3. Fringe patterns of (a) square burner; (b) rectangular burner. Temperatures along reference line
for (c) square burner; (d) rectangular burner.

The present study demonstrates the possibility of applying MachZehnder Interferometry and
digital photography to real time study of the temperature eld in a premixed butane/air slot
open ame jet. In comparison with the conventional method of employing thermocouples to
take temperature eld data, it has the obvious advantages of being non-contact, non-intrusive,
no compensation required for radiation and Thermo-Chemical-Heat-Release (TCHR) efect
and being able to obtain whole eld data at the same time. It can be employed to study the
temperature eld of both steady ames and transient ames. It is fore-seen that the method
can be adapted or further enhanced to study many combustion phenomena in the future. [1]

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 16

Methods of measuring flame temperature

CHAPTER 3

CONCLUSION
The methods which we are using for measurement of temperature of the flame giving close
results to the actual because of latest computerized digital measurement system helps to
provide errorless results. So that the results obtained will help better in design of burners and
flame analysis. But the main drawback of advanced measurement system is the cost of the
equipment & repeatability, because the instruments which are employed in the measurement
are much higher than conventional thermocouple measurement system and skill required
handling the measurement system. So in future we can expect efficient and economic system
will be developed to measure the flame temperature very accurately.

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 17

Methods of measuring flame temperature

CHAPTER 4

REFERENCES
1. Paper on Temperature eld measurement of a premixed butane/air slot laminar ame
jet with MachZehnder Interferometry by J.A. Qi, W.O. Wong, C.W. Leung *,
D.W. Yuen Department of Mechanical Engineering, The Hong Kong Polytechnic
University, 1 Yuk Choi Road, Hung Hom, Kowloon, Hong Kong.
2. Paper on Problems Encountered in Fluctuating Flame Temperature Measurements by
Thermocouple Nadir Yilmaz,Walt Gill ,A. Burl Donaldson and Ralph E. Lucero.
3. Paper on Flame structure studies: past, present and future by Robert M. Fristrom
The Johns Hopkins University Applied Physics Laboratory Laurel, Maryland
20707, USA
4. Paper on Temperature measurement using a fine wire compensated thermometry by
BOULAHLIB M.S & AMARA I. Department of Mechanical Engineering,
University Mentouri Constantine,25017 Constantine, ALGERIA

MTECH THERMAL ENGINEERING NITK SURATHKAL

Page 18