VOLUME 30

ISSUE 2
o f A c h ie v e m e n t s in M a t e r ia ls October
a n d M a n u f a c t u r in g E n g in e e r in g 2008

Problems for det ermining t he t hermal

conduct ivit y of TBCs by laser -flash met hod
O. Alt un a,* , Y. Er han Boke b, A. Kalemt as c
a Mechanical Engineering Deparment, Eskisehir Osmangazi University,
Batimeselik, Eskisehir, Turkey
b Mechanical Engineering Deparment, Istanbul Technical University,
Gumussuyu, Istanbul, Turkey
c Materials Science and Engineering Deparment, Anadolu University,
Eskisehir, Turkey
* Corresponding author: E-mail address: [email protected]
Received 13.06.2008; published in revised form 01.10.2008

Mat er ials

Ab s t r Ac t
Pur pose: The purpose of this paper was to investigate the parameters which effect the results of determining the
thermal conductivity of thermal barrier coatings (TBCs) by laser-flash method.
Design/methodology/approach: The air plasma-spray (APS) technique was used to deposition of two- and
three-layered samples. Two-layered samples were composed of metal substrate (321 stainless steel), and ceramic
top coat (8YSZ). Three-layered samples were composed of metal substrate (321 stainless steel), bond coat
(NiCrAlY) and top coat (8YSZ). Thermal diffusivity of each layer have been measured in the temperature range
from room temperature (RT) to 900C by laser-flash method. The thermal conductivity was calculated with
respect to density, specific heat and diffusivity of the materials.
Findings: Obtained results show that the specific heat, density and thicknesses of metal substrate, bond coat and
top coat play important role in the thermal conductivity measurement.
Research limitations/implications: To obtain the correct results in laser-flash technique thickness, density, and
cp of the materials are needed to be measured accurately and surface smoothness of samples should be provided
sensitively. Errors in these parameters cause high deviations in measurements.
Pr actical implications: It has been aimed offer an insight into the experimental determination of thermal
conductivity of layered TBC system which are used in high technologic applications.
Or iginality/value: Laser-flash method is the most widely used experimental technique to determine the
thermal conductivity of APS TBCs at high temperatures. The research contributes to better understanding and
recognition the importance of sample preparation in laser-flash method.
Keywor ds: Ceramics; Thermal conductivity; TBCs; Laser-flash method

thermal barriers TBC deposited upon surfaces of super-alloy

1..Intr
1 I n oduction
odu io n surfaces, operating under load in high temperatures [1].
TBCs have been widely used in hot-section metal components in
Oxide ceramic materials feature high melting temperature,
gas turbines either to increase the inlet temperature with a consequent
relatively low density and low thermal conductivity as well as
improvement to the efficiency or to reduce the requirements for the
considerable corrosion resistance to chemically active media in high
cooling air [2-5]. The typical TBC used in gas turbines consists of a
temperatures. Ceramic coatings, however, seem more effective as
bond coat produced by the vacuum or low pressure plasma-sprayed

Copyright by International OCSCO World Press. All rights reserved. 2008 R es ear c h p ap er 115
 Journal of Achievements in Materials and Manufacturing Engineering Volume 30 Issue 2 October 2008

MCrAlY (M = Ni, Co) and a top coat of yttria partially stabilized It is essential to know the thermal diffusivity value of each
zirconia made by the atmospheric plasma spraying or electron beam- layer for layered systems in laser-flash method [18-20]. Hence,
physical vapor deposition (EB-PVD) [6-7]. diffusivity of 321 stainless steel was measured first. Then for the
Within the last decade there has been a consistent growth in samples composed of substrate and bond coat, thermal diffusivity
the use of composite ceramic materials and layer-coated ceramics. of the bond coat was recorded. Finally diffusivity of Air Plasma
Excellent chemical and wear resistance, a wide range of electrical Sprayed (APS) coating was measured. The specific heat of each
and thermal properties, and high service temperatures have made layer was determined for the purpose of calculating thermal
these materials extremely valuable to industry. Due to their conductivity in view of past literature data.
increasing importance, and the wide range of shapes, sizes, and
composition in which these materials are produced, it is essential
to have reliable methods available to measure their thermal 2.1.
2 .1 .Lazer
L a z eflash
f l a technique
h e h n iq u e
conductivity and diffusivity [8].
There are a variety absolute and comparative methods The heat balance equation for transient conditions may be
currently used to measure thermal conductivity or diffusivity, written as
including both transient and steady-state [8-10] techniques. Laser-
flash [11-13], hot-wire, and three-omega methods are the most dT (1)
widely used transient techniques [14]. The laser flash method is T intrenal source/sink Cp
dt
used commonly and typical of the non-contact transient method.
An instantaneous heat pulse is generated by the absorption of the If there are no internal source:
laser energy on the front surface of the sample and it transmits to
the rear surface on which the temperature rise is detected by an IR dT (2)
sensor [15]. This method is widely used because of the small T Cp
sample size required and the ease and rapidity with which dt
measurement can be taken. The disadvantage of this method is
that density and heat capacity also need to be known or measured For homogeneous materials whose thermal conductivity is nearly
to calculate thermal conductivity. In addition, the use of indirect independent of temperature, we may treat as a constant. Then
2
transient methods may yield questionable results in complex T becomes T , and Eq. (2) can be written as:
composites or layered ceramics where there are unusual patterns
of heat flow within the body of the material [8]. 2 dT (3)
T CP
The thermal conductivity of plasma sprayed zirconia dt
(stabilized with 7, 8, and 20 wt % yttria) and of NiAl, NiCr,
NiCrAl, NiCrAlY and NiCoCrAlY bond coatings have been or
measured in the temperature range 27 C 727 C with the use of 2 1 dT (4)
T
laser-flash method [16]. dt
Plasma sprayed coatings have been investigated between the
room temperature and 1200 C [17] as an example of Al2O3 and where / Cp is the thermal diffusivity. For one-dimensional
ZrO2 based coatings.
This method was extended to the numerical approach was heat flow,
chosen for the determination of contact resistance, which
considers finite pulse time, heat loss, and transparency [18]. d 2T dT
Hohenauer and Vozr [19] have been presented measurements of dx 2 dt
the thermal diffusivity on two- and three-layered copper/alumina
composites using the laser-flash method. For heterogeneous materials, is not independent of
In this study the thermal conductivity of two-layered samples
position and should not be moved from behind the gradient
which are composed of substrate, ceramic coating and three-
operator (Eq.(2) and Eq.(3)). In principle, then, the concept of
layered samples which are composed of substrate, bond coat,
diffusivity is inapplicable to heterogeneous materials. However,
ceramic coating have been measured by laser-flash method.
in practice, materials are never homogeneous, as point defects,
The parameters which will effect the measurement results in
dislocations, grain boundaries, voids and so on are present even in
determining the thermal conductivity of air plasma sprayed TBC
so-called homogeneous materials. Yet diffusivity techniques have
coating (APS) by laser-flash method and will increase the
been applied to these materials for many years. Successful
sensitivity of measurement and the reasons of errors in results
attemps to extend diffusivity techniques to obviously
were disscussed.
heterogenous materials have been made for many years. These
efforts have been intensified since the 1960s owing to the
development of the flash method of measuring diffusivity. This
2. .Mater
2 M a eialia
and
l a nmethod
d m e hod method uses relatively simple sample geometry, permits rapid
measurements, and it can handle materials with a large range of
In this research, as type of the coating using yttria stabilized diffusivity values over extended temperature intervals with a
zirconia 8 wt % (8YSZ) air plasma sprayed thermal barrier single apparatus [17].
coating was analyzed.

116 R es ear c h p ap er O. Altun, Y. Erhan Boke, A. Kalemtas

 Mat er ials

The laser flash method, ASTM E1461-02 [21], is shown pulse lenght of 0.2-1.2 ms. The temperature rise of the rear side of
schematically in Figure 1. The samples front surface recevies a the sample is recorded without contact by means of an indium
pulse of energy from the laser, which soon raises the back face antimonide infrared detector. Before the measurement, the coated
temperature a degree or two. The rear face temperature response surface of the samples were coated with a thin layer of gold. This
is normalized and compared with the theoretical model based on thin golden layer prevents the direct transition of the laser beams
Carslaw and Jaegers solution to one-dimensional heat flow. A and it helps the energy transfer to the sample. All samples (two
solition of a one-dimensional thermal diffusion equation with an
and three-layered samples) were graphite coated for absorption
initial condition of a heat pulse is,
and emission.
(5) Furthermore, using the programs that take the radial heat loss
Q n2 2
t into acount, calculations were made.
T (d , t ) 1 2 ( 1) n exp
Cpd r 2 n 1 d 2

3 .D i u io n o f h e
where Q is the absorbed energy at the front surface, and d and r 3. Discussion
e x p e im eofn the
a l exper
e uimental
l r esults
are the thickness and the radius of the sample, respectively. The
thermal diffusivity is deduced from the time to reach one half
The thermal conductivity of samples were calculated from the
of the maximum value according to the formula,
thermal diffusivity, specific heat and density data with the aid of
equation (7).
1.37d 2 (6)
2
t1/ 2
Cp . . (7)
where d is the sample thickness and t 1 / 2 is the elapsed time
where Cp specific heat, density and thermal diffusivity.
needed for the rear face temperature to reach one-half of its
maximum rise (Figure1) [17].
321 Stainless Steel: Small disc-shaped samples of 321 stainless
steel which 12.7 mm in diameter with a thickness of 4.21 mm
have been used. The density was measured by the Archimedes
method and evaluated 7.97 g/cm3. Specific heat values of steel
were obtained from literature [22] and properties of the samples
are given in Table 1.

Table 1.
321 Stainless steel sample
Sample thickness Density Specific Heat
(mm) (g/cm3) (J/g.K)
4.21 7.970 0.5

For the thermal diffusivity measurements both sides of the

samples were coated carbon. Carbon increases the absorption and
Fig. 1. Schematic of flash diffusivity experimental setup the emission. This measurements were done in the temperatue
range 23 C 900 C and in Nitrogen gas (N2) atmosphere at a
flow rate of 100 mbar/s using the laser-flash method (Netzsch,
2.2.
2 .2 Samples
. s a m p le LFA 457, Selb, Germany). The thermal conductivity was
calculated from the thermal diffusivity, density, and specific heat
Both two- and three-layered samples were deposited, using data with the aid of equation (7). Obtained diffusivity data of steel
the same direct-current 9MB plasma torch (Sulzer Metco) on 321 were used in diffusivity measurement of bond coat and 8YSZ
stainless steel disc substrates (diameter 12.7 mm.) coating.
Two-layered samples are composed of substrate (321 stainless The experimental results of the thermal conductivity of 321
steel) and plasma sprayed zirconia coating (8YSZ), and three- stainless steel is plotted in Figure 2. Thermal conductivity of
layered samples are composed of substrate, bond coat (NiCrAlY) stainless steel is agreement of metal characteristic. It is observed
and plasma sprayed coating. that thermal conductivity of steel inreases by temperature.
The measurements were performed with the laser-flash unit The thermal conductivity of the steel at 100 C is about ~ 15.1
LFA 457. Measurements were done between the room W/mK and at 500 C is about ~18.3 W/mK as obtained from the
temperature and 900 C in Nitrogen (N2) atmosphere and were laser-falsh method. In this temperature this values are 16.1 W/mK
done at least three times and average values were taken. The laser and 22.2 W/mK respectively in literature. A maximum of error
is a pulsed Nd-YAG type with a wavelenght of 1064 m and a 11 % was detected.

Problems for determining the thermal conductivity of TBCs by laser-flash method 117
 Journal of Achievements in Materials and Manufacturing Engineering Volume 30 Issue 2 October 2008

25 25

20 20
k (W/mK)

k (W/mK)
15 15

10 10
LFA 457-Group1
5 5 Brandt and Pawlowski (1986)

0 0
0 200 400 600 800 1000 0 200 400 600 800 1000
T (C) T( C)

Fig. 2. Thermal conductivity of 321 stainless steel Fig. 3. Thermal conductivity of plasma sprayed NiCrAlY

Nichel-chromium + aluminium, yttrium (NiCrAlY) bond coat: In

The thermal conductivity data of the coatings are plotted in
these measurements three samples which have same properties
have been used. The samples were composed of substrate (321 Figure 3. The conductivity of the NiCrAlY sample sprayed
stainless steel) and bond coat (NiCrAlY). The density of the bond measured by Brandt et al. [16] is shown for comparison. A
coat measurement by the Archimedes method and apparent maximum of errors 7% is detected by this figure.
densitywas found about 3.38 gr/cm3. The properties of the Thermal conductivity of typical bond coat for thermal barrier
samples are given in Table 2. Specific heat of bond coat are coatings was measured. The thermal conductivity of NiCrAlY
obtained from MTDATA (version 4.74) source program. increases from 10.494 to 23.327 W/mK in the temperature range
23-900 C. Between 23-300 C and 500-900 C, the thermal
Table 2. conductivity clearly increases. This values are close to the
NiCrAlY coatings literature values.
Group 1 The conductivity of Brandt measurement at 400 C is about ~
Sample Sample Sample 16.9 W/mK and 900 C is about ~ 22.5 W/mK as obtained from
Z1A Z1B Z1C laser-flash method. In experiments at same temperatures 17.1
Steel thickness (mm) 4.27 4.26 4.51 W/mK and 23.3 W/mK are obtained respectively.
Coating thickness
No No No Yttria-stabilized zirconia plasma sprayed coatings (8YSZ):
(mm)
NiCrAlY bond coating 0.2 0.21 0.23 Thermal conductivity of 8YSZ was measured two different form.
First, three-layered samples which composed substrate, bond coat
Density (g/cm3) 3.376 3.376 3.376
and 8YSZ coatings was measured. The coatings have a density of
5.0 gr/cm3. Specific heat of zirconia are obtained from MTDATA
The thermal diffusivity ( ) was determined from equation (6). (version 4.74) source program. The properties of three-layered
Obtained values were used in diffusivity measurement of three- samples are given in Table 3.
layered samples.

Table 3.
Three-layered plasma-sprayed zirconia coatings samples
Group 3 Group 4 Group 5
Z2A Z3B Z3C Z5A Z5B Z5C Z6A Z6B Z6C

Steel thickness/mm 4.28 4.19 4.37 4.34 4.24 4.26 4.21 4.33 4.21

Coating thickness/mm 0.5 0.59 0.57 0.72 0.85 0.66 0.46 0.54 0.75

NiCrAlY bond coating/ mm 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21 0.21

Avarege coatings thickness/mm 0.55 0.74 0.58

118 R es ear c h p ap er O. Altun, Y. Erhan Boke, A. Kalemtas

 Mat er ials

Two- and three-layered measurement were performed between Table 4.

RT and 900 C using the laser-flash method. Both sides of the Two-layered plasma-sprayed zirconia coatings samples
samples were coated with gold and carbon. Gold prevents the Group 2
direct transmission of the beam and improves energy transfer to Sample Sample Sample
the samples. Measurements were carried out under a nitrogen gas Z2A Z2B Z2C
atmosphere. The thermal conductivity of 8YSZ coatings were Steel thickness
calculated according to equation (7). 3.18 3.14 3.13
(mm)
Coating thickness
0.22 0.46 0.27
(mm)
9 NiCrAlY bond
NO NO NO
8 coating
7
Secondly two-layered samples which composed substrate and
6 8YSZ coatings was measured. The properties of two-layered
k (W/mK)

5 samples are given Table 4. Specific heat of zirconia for two-layer

sample are obtained from literature [16,23]
4
The experimental results of thermal conductivity for two-
3 layered samples are plotted in Figure 5. This figure is produced by
Group3 averaging the data obtained for the three samples in Table 4. In
2 Group4 this measurement, the experimental thermal conductivity of 8YSZ
1 Group5 was obtained at RT about ~7.7 W/mK and 900 C about
0 ~ 12.12 W/mK. When these values were compared to the values
0 200 400 600 800 1000
which in literature it was seen that the values were ten fold bigger.
However for three-layer measurements these values were 4 to 8
T ( C) fold bigger. Cp values used in three-layer measurements provides
Fig. 4. Thermal conductivity of three-layered 8YSZ coatings better results than cp values which in literature.

14
4. .Conclusion
4 c o n lu io n
12
The thermal conductivity of 8YSZ are not strongly
10 temperature dependent. Reliable thermal conductivity values for
k (W/mK)

8 TBCs can be obtained on two- and three-layer composites by

laser-flash technique under specified conditions.
6 The determination of the thermal conductivity of a two- and
three-layered systems is a dependent measurement. An estimation
4 of the thermal conductivity of one layer requires in addition
2 Group2
knowledge of other relevant properties (thermal diffusivity,
specific heat, density, and thickness of layers) in order to
0 determine the thermal conductivity of the remaining layers. Errors
0 200 400 600 800 1000 in the measurement of these additional parameters are propagated
through the data reduction and results in error in the thermal
T ( C)
conductivity determination. Especially the thickness of the
coating is very effective on the thermal conductivity (Eq.2). For
Fig. 5. Thermal conductivity of two-layer 8YSZ coatings example a 10% error in coating thickness, causes a 20% error in
the calculated conductivity values. Additionally the top and
bottom surface of samples should be paralell to each other and
Figure 4. represents the temperature dependence of the being homogeneous in all layer thicknesses play important role in
thermal conductivity of three-layered samples, containing measurement accuracy.
different coating thicknesses. In this figure, for three group Furthermore, another parameter that affects measurement
samples with different thicknesses the thermal conductivity was sensitivity, is surface smoothness of both metal and coating. And
evaluated neglecting the radiation influence. Up to 300 C thermal another trouble met in measurements is because zirconia partially
conductivity of all three group measurements are equal. At higher transparent at high temperatures. It makes difficult the
temperature the thicker samples (Group 5) show an apparently measurement of surface temperature of zirconia. In that reason
higher value of thermal conductivity. The heat pulse due to samples were gold coated for energy transfers to the sample and
radiation reaches the sample rear side instantaneously whereas the graphite coated for absorption and emission.
diffusion takes some time. The time difference is less for thin When the results compared to the literature, it was seen that
sample. This means a lower influence on the temperature curve approximately ten fold bigger values were obtained. Thinking that
for the thinner samples. the errors are the results of measurements done as three layer, two

Problems for determining the thermal conductivity of TBCs by laser-flash method 119
 Journal of Achievements in Materials and Manufacturing Engineering Volume 30 Issue 2 October 2008

layer measurements were done. It is assessed that this error [10] A.J. Slifka, Thermal conductivity apparatus for steady-state,
caused from that the coating thickness is not homogeneous, and comparative measurement of ceramic coatings, Journal of
Cp values are obtained from literature. Research of the National Institute of Standards and
Therefore to determine the thermal conductivity of ceramic Technology 105 (2001) 591-605.
coating properly, Cp, density and thicknesses of metal substrate [11] W.J. Parker, R.J. Jenkins, C.P. Butler, and G.L. Abbott,
and coating are needed to be measured accurately. Additionally it 1961, Flash method of determining thermal diffusivity, heat
is required that the surface smoothness should be lower than 1% capacity, and thermal conductivity, Journal of Applied
of thickness. Errors in these parameters cause significant Physics 32 (1961) 1679-1684.
deviations for measurement of the thermal conductivity of TBCs. [12] R.D. Cowan, Pulse method of measuring thermal diffusivity at
high temperatures, Journal of Applied Physics 34 (1962) 926-927.
[13] J.A. Cape, and G.W. Lehman, Temperature and finite pulse-
time effects in the flash method for measuring thermal
Refer
r e f eences
en e diffusivity, Journal of Applied Physics 34 (1963) 1909-1913.
[14] T.A. Taylor, and R.E. Taylor, Testing of stability and
[1] K. Kobyla skaSzkaradek, Thermal barrier ZrO2-Y2O3 thermal properties of thermal barrier coatings, ASM
obtained by plasma spraying method and laser melting, Handbook 5 (1999) 654-659.
Journal of Achievements in Materials and Manufacturing [15] M. Sorai, Comprehensive handbook of calorimetry and
Engineering 17 (2006) 77-80. thermal analysis, John Wiley & Sons, Ltd, (2004) 172.
[2] K.A. Khor, S. Jana, Pulse laser processing of plasma [16] R. Brandt, L. Pawlowski, and G. Neuer, Specific heat and
sprayed thermal barrier coating, Journal of Materials thermal conductivity of plasma sprayed yttria-stabilized zirconia
Processing Technology 66 (1996) 4-8. and NiAl, NiCr, NiCrAl, NiCrAlY, NiCoCrAlY Coatings, High
[3] M. Konter, M. Thumann, Materials and manufacturing of Temperatures -High Pressures 18 (1986) 65-77.
advanced industrial gas turbine components, Journal of [17] R.E. Taylor, Thermal conductivity determination of thermal
Materials Processing Technology 92-117 (2001) 386-390. barrier coatings, Materials Science and Engineering A245
[4] J.F. Li, H.L. Liao, C.X. Ding, C. Coddet, Optimizing the (1998) 160-167.
plasma spray process parameters of yttria stabilized zirconia [18] F. Schmitz, D. Hehn, H.R. Mainer, Evaluation of laser-flash
coatings using a uniform design of experiments, Journal of measurements by means of numerical solution of the heat
Materials Processing Technology 160 (2005) 34-42. conduction equation, High Temperature - High Pressures 31
[5] L. Swad ba, G. Moskal, B. Mendala, T. Gancarczyk, (1998) 203-211.
Characterisation of air plasma sprayed TBC coating during [19] W. Hohenauer, L. Vozr, An Estimation of thermophysical
isothermal oxidation at 1100 C, Journal of Achievements in properties of layered materials by the laser-flash method,
Materials and Manufacturing Engineering 21 (2007) 81-84. High Temperature - High Pressures 33 (2001) 17-25.
[6] G. Moskal, The porosity assessment of thermal barrier coatings [20] R.L. Mcmasters, R.B. Dinwiddie, A. Haji-Sheikh, Estima-
obtained by APS method, Journal of Achievements in Materials ting the thermal conductivity of a film on a known substrate,
and Manufacturing Engineering 20 (2007) 483-486. Journal of Thermophysics and Heat Transfer 21-4 (2007)
[7] D. Stover, C. Funke, Directions of the development of 681-687.
thermal barrier coatings in energy applications, Journal of [21] ASTM E1461, Standard test method for thermal diffusivity
Materials Processing Technology 92-93 (1999) 195-202. by the flash method, ASTM Standards, American Society
[8] B.J. Filla, A Steady-state high-temperature apparatus for for Testing and Materials-Philadelphia, 2002.
measuring thermal conductivity of ceramics, Review of [22] ASM Handbook, Properties and selection: irons, steels and
Scientific Instruments 68-7 (1997) 2822-2829. high performance alloys 1, 1986.
[9] K. An, M. K Han., An experimental technique to evaluate [23] R.E. Taylor, X. Wang, X. Xu, Thermophysical properties of
the effective thermal conductivity of Y2O3 stabilized ZrO2 thermal barrier coatings, Surface & Coatings Technology.
coatings, Journal of Materials Science 41 (2006) 2113-2120. 120-121 (1999) 89-95.

120 R es ear c h p ap er READING DIRECT: www.journalamme.org