Journal of Alloys and Compounds 489 (2010) 482–487

Contents lists available at ScienceDirect

Journal of Alloys and Compounds

journal homepage: www.elsevier.com/locate/jallcom

Emissivity measurements on aeronautical alloys

L. del Campo a,∗ , R.B. Pérez-Sáez a,b,∗∗ , L. González-Fernández a , X. Esquisabel c ,
I. Fernández c , P. González-Martín d , M.J. Tello a,b
a
Departamento de Física de la Materia Condensada, Facultad de Ciencia y Tecnología, Universidad del País Vasco, Barrio Sarriena s/n, 48940 Leioa, Bizkaia, Spain
b
Instituto de Síntesis y Estudio de Materiales, Universidad del País Vasco, Apdo. 644, 48080 Bilbao, Spain
c
Industria de Turbo Propulsores, S.A., Planta de Zamudio, Edificio 300, 48170 Zamudio, Bizkaia, Spain
d
Industria de Turbo Propulsores, S.A., Parque empresarial San Fernando, Avda. Castilla 2, 28830 San Fernando de Henares, Madrid, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The emissivity of three Ni and Co based aeronautical alloys is analyzed in this paper. These alloys are
Received 16 July 2009 employed in high temperature environments whenever good corrosion resistance, high temperature
Received in revised form resistance and high strength are essential. Thus, apart from the aeronautical industry, these alloys are
15 September 2009
also used in other technological applications, as for example, aerospace, nuclear reactors, and tooling.
Accepted 16 September 2009
The results in this paper extend the emissivity data for these alloys available in the literature. Emissivity
Available online 25 September 2009
dependence on the radiation wavelength (2–22 ␮m), sample temperature (200–650 ◦ C) and emission
angle (0–85◦ ) has been investigated. In addition, the effect of surface finish and oxidation has also been
Keywords:
High-temperature alloys
taken into consideration. The data in this paper have several applications, as temperature measurement
Thermodynamic properties of a target by pyrometry, low observability of airplanes and thermal radiation heat transfer simulation
Infrared emissivity in airplane nozzles or furnaces.
Infrared spectroscopy © 2009 Elsevier B.V. All rights reserved.

1. Introduction perature fields inside the nozzle, low observability of the airplanes
to prevent missile detection, etc.
The emissivity is a property which reveals how much radiation a In order to obtain reliable emissivity data for aeronautical appli-
given body emits as compared to a blackbody [1]. This magnitude is cations, three alloys are analyzed in this paper: Inconel 718, René 41
necessary in quite a lot of industrial and technological applications, and Haynes 25. These Ni and Co based alloys are not only employed
as for example when taking into account thermal radiation heat in the aeronautical industry. They are used in high temperature
transfer, especially in high temperature and vacuum environments. environments, whenever good oxidation and corrosion resistance
It is also essential when measuring the temperature of a target and high mechanical strength are essential; for example, aerospace,
by radiometry (pyrometry). The emissivity is highly influenced by nuclear reactors, and tooling applications. Inconel 718 is even uti-
the surface state, thus it depends on surface roughness, oxidation, lized for cryogenic uses, as cryogenic storage tanks. The emissivity
machining process, etc. Due to this influence, it is advisable to mea- values can be also very usable in all these applications.
sure the emissivity of a sample in the operating conditions. Ideally the emissivity should be measured under the service
In the last years, there is an increasing requirement of the conditions of the required application. If this is not possible
infrared emissivity of the alloys used in the aeronautical indus- one should select adequate emissivity values from literature.
try. The characterization of the emissivity of these alloys is crucial Therefore the surface properties of the specimen have to be well-
for several purposes, as for example, temperature measurement characterized as they influence the emissivity significantly. Any
by pyrometry inside the airplane nozzle where a contact sensor is reference has been found where emissivity data for René 41 or
unfeasible, thermal radiation heat transfer in order to simulate tem- Haynes 25 has been published. There are some references where
the emissivity of the Inconel 718 has been measured [2–5], but
only the total hemispherical emissivity or the spectral emissivity
at a particular wavelength are given. Thus, Refs. [2,3] show that the
∗ Corresponding author. Present address: Conditions Extrêmes et Matériaux: total hemispherical emissivity rises with increasing surface rough-
Haute Température et Irradiation (CEMHTI), Centre National de la Recherche Scien- ness, temperature and oxidation state. Refs. [4,5] give some spectral
tifique (CNRS), Site Haute Température, 1D Av. de la Recherche Scientifique, 45071 emissivity values at 684.5 nm and 1.6 ␮m, respectively.
Orléans Cedex 2, France. Tel.: +33 02 38 25 56 70; fax: +33 02 38 63 81 03.
∗∗ Corresponding author. Tel: +34 94 601 2655; fax: +34 94 601 3500. Some other references have been found for similar Ni and Co
E-mail addresses: [email protected] (L. del Campo), based alloys. For example, Ref. [6] shows indirect emissivity mea-
[email protected] (R.B. Pérez-Sáez). surements (calculated using reflectivity data) for some Ni and Co

0925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2009.09.091
 L. del Campo et al. / Journal of Alloys and Compounds 489 (2010) 482–487 483

based alloys. Refs. [7,8] are related to a Thermophysical properties tor. A PID temperature controller is used to control the sample temperature. The
database where emissivity results for several compounds are found, apparatus permits to obtain the angle dependence of the emissivity by tilting the
sample and changing its polar angle (angle between the emission direction and
for example Inconel 600, Inconel 601, and Inconel 625. Finally, some
the normal to the surface). The radiation emitted by the sample is detected by
emissivity data for Inconel 600, Inconel X and Haynes 230 are given a Fourier transform infrared (FT-IR) spectrometer (Bruker IFS66v/S), and its radi-
in Refs. [9–11]. ance is compared with the radiance of a blackbody radiator. A simple comparison
In this paper, the available information for Inconel 718 is is not advisable because background radiation in the sample chamber and offset
extended, and the first emissivity data for René 41 and Haynes 25 radiation inside the spectrometer have to be taken into account. Therefore, a shut-
ter is placed in front of the blackbody, and a third measurement is performed.
are presented. Emissivity dependence on the radiation wavelength Finally, the three signals are used to get the sample spectral emissivity (ε) according
(2–22 ␮m), sample temperature (200–650 ◦ C) and emission angle to:
(0–85◦ ) has been investigated. Additionally, the effect of surface
finish and oxidation has also been taken into consideration. The FT [Is − Ish ] Lbb − εsh Lsh εsh Lsh − Lsur
ε= × + (1)
FT [Ibb − Ish ] Ls − Lsur Ls − Lsur
most widespread conditions have been accounted for in order to
characterize the emissivity of these alloys the best possible. where the subindices s, bb, sh and sur have been used for sample, blackbody, shutter
and sample surroundings respectively. In the equation, FT means Fourier transform,
2. Experimental Ii is used for i’s interferogram, and Li for the blackbody radiance (Planck’s equation)
at temperature Ti . The measurement method is more precisely described in Refs.
2.1. Samples [12,13].
Before the emissivity measurements are performed, all the samples have been
Three Ni and Co based alloys have been investigated: Inconel 718, René 41, and cleaned with acetone in an ultrasonic bath. Two K-type thermocouples have been
Haynes 25. Table 1 shows the components of the three alloys in weight percentage. spot welded in each sample, at two symmetric points 5 mm away from the centre.
As seen, the Inconel 718 and René 41 are Ni based, whereas the Haynes is Co based. They do not disturb the radiation measurement as the area focalized at the detector
It is remarkable that the three of them have a high Cr content. is a 5 mm diameter disk centred at the sample. Before the samples are heated, the
Disk shaped samples, 2–3 mm thick cut from a 60 mm diameter rod have been sample chamber is evacuated and a slightly reducing gas (N2 + 5%H2 ) is introduced.
used. The samples have been wire-cut by electrical discharge machining (EDM). The H2 is used to minimize sample oxidation. Thus, when the atmosphere inside
Some of them have been additionally brushed, and certain also sandblasted. Thus, the chamber is ready, the sample is heated till the desired temperature is achieved.
three surface finishes have been analyzed: wire-cut EDM, brushed, and sandblasted. Once the sample temperature is stable, the sample spectrum is acquired, next, the
The wire-cut EDM surfaces are dark coloured and look inhomogeneous due to a blackbody signal is measured, and then a third spectrum is obtained with the shutter
recast layer that is formed on the surface during the machining process. The brush- placed in front of the blackbody. Finally, Eq. (1) is used to find the emissivity. The
ing removes the recast layer, and smoothes the sample surface. This machining estimated uncertainty is less than 5%.
creates a surface profile with a strong unidirectional lay pattern. Finally, some of the
brushed surfaces have been additionally sandblasted. This surface finish produces
an isotropic surface profile.
Table 2 shows some roughness properties of the nine types of samples analyzed.
3. Results and discussion
The surface roughness average (Ra ), average maximum height (Rz ) and maximum
height of the profile (Rt ) have been measured for one sample of each type. As seen, the The spectral normal emissivity has been measured for all the
roughest samples are the wire-cut EDM. On the other hand, it is remarkable that the sample types shown in Table 2, for temperatures ranging from 200
brushing and the sandblasting have produced very dissimilar surface topographies
to 650 ◦ C, and for wavelengths between 2 and 22 ␮m. Two heat-
but with very similar roughness average parameters.
ing cycles have been measured for each sample. Furthermore, the
2.2. Emissivity measurements spectral directional emissivity has been measured for sandblasted
samples as a function of the emission angle, between the normal
The emissivity of the samples has been measured by means of a highly accu- to the surface and 85◦ . Between 85◦ and 90◦ the emissivity can be
rate homemade radiometer [12]. The samples are introduced in a vacuum sample
chamber, and a resistance heater is used to heat them. The temperature of the
extrapolated so that the hemispherical emissivity can be obtained.
sample surface is measured by bare K-type spot welded thermocouples, which Additionally, the oxidation effect on the emissivity has also been
are placed on the sample surface, out of the area viewed by the infrared detec- studied for brushed samples.

Table 1
Chemical composition of Inconel 718, René 41 and Haynes 25 in weight percentage.

Al B C Co Cr Cu Fe Mn

Inconel 718 0.2–0.8 Max. 0.006 Max. 0.08 Max. 1 17–21 Max. 0.3 Balance (∼20) Max. 0.35
René 41 1.4–1.8 0.003–0.01 0.04–0.12 10–12 17.5–20 Max. 5 Max. 0.1
Haynes 25 0.05–0.15 Balance (∼50) 19–21 Max. 3 1–2

Mo Nb Ni P S Si Ti W

Inconel 718 2.8–3.3 4.75–5.5 50–55 Max. 0.015 Max. 0.015 Max. 0.35 0.65–1.15
René 41 9–10.5 Balance (∼50) Max. 0.015 Max. 0.5 3–3.3
Haynes 25 9–11 Max. 0.4 14–16

Table 2
Summary of the samples with their surface roughness: roughness average (Ra ), average maximum height (Rz ), and maximum height of the profile (Rt ).

Sample type Alloy Surface finish Ra (␮m) Rz (␮m) Rt (␮m)

1 Inconel 718 Brushed 1.3 8.2 9.9

2 Inconel 718 Sandblasted 1.6 8.0 –
3 Inconel 718 Wire-cut EDM 2.4 14.7 18.6
4 René 41 Brushed 1.5 10.1 13.0
5 René 41 Sandblasted 1.7 8.3 –
6 René 41 Wire-cut EDM 2.4 14.5 20.1
7 Haynes 25 Brushed 1.2 7.6 10.7
8 Haynes 25 Sandblasted 1.6 7.8 –
9 Haynes 25 Wire-cut EDM 2.7 16.4 19.9
484 L. del Campo et al. / Journal of Alloys and Compounds 489 (2010) 482–487

Fig. 1. Spectral normal emissivity (ε) of brushed samples as a function the wave- Fig. 2. Spectral normal emissivity (ε) of sandblasted samples as a function the wave-
length () for several temperatures: (a) Inconel 718 (sample type 1), (b) René 41 length () for several temperatures: (a) Inconel 718 (sample type 2), (b) René 41
(sample type 4), and (c) Haynes 25 (sample type 7). (sample type 5), and (c) Haynes 25 (sample type 8).

3.1. Dependence on temperature and wavelength emissivity obtained for brushed and sandblasted samples is the
same. The same emissivity increase is observed at short wave-
In Fig. 1, the measured spectral normal emissivity of brushed lengths, and accordingly, the sandblasted samples also look bluish
samples is shown as a function of the wavelength, for several after heating.
sample temperatures. As seen, the temperature and wavelength In Fig. 3 the same emissivity spectra are shown for wire-cut EDM
dependence of the emissivity for the three brushed alloys is nearly surfaces. The recast layer contain some compounds that evaporate
the same, that is, it decreases with increasing wavelength, and a at high temperature and for that reason the emissivity varies during
slight increase of the emissivity with temperature is observed at the first heating cycle, thus the results obtained during the second
short and long wavelengths. heating cycle are shown. As seen, the spectral dependence is basi-
Nevertheless, the emissivity increase at short and long wave- cally the same as for brushed and sandblasted samples. However,
lengths has not the same cause, i.e., the emissivity at long the emissivity does not increase with temperature at short wave-
wavelength returns to its initial value when cooling down, while at length. This is probably due to the recast layer formed on the surface
short wavelength, the increase in emissivity observed during the during the machining process, which inhibits further oxidation of
first cycle in maintained, that is, the emissivity does not return to the samples.
its initial value when cooling down. Thus, the emissivity increase For the sake of a better visualisation of the variation of the spec-
at long wavelengths is really related to a change in the emissive tral normal emissivity with the temperature, in Fig. 4 the emissivity
behaviour of the sample as the temperature changes, while the of the nine types of samples is shown as a function of the tem-
increase of the emissivity at short wavelength is related to a change perature for several wavelengths (2.5 ␮m (Fig. 4a), 10 ␮m (Fig. 4b)
in the sample surface during heating. As explained, the measure- and 20 ␮m (Fig. 4c)). For metallic samples, the emissivity usually
ments have been performed in a slightly reducing atmosphere. increases with increasing temperature, but for these alloys, the
Nevertheless, it is known that these Ni and Co alloys, as well as some emissivity nearly does not depend on the temperature for a broad
stainless steels oxidize to some extent in reducing atmospheres. spectral range in the studied temperature region.
Actually, these alloys have a high Cr content, which oxidizes even
in reducing atmospheres, as long as any small amount of oxygen 3.2. Effect of surface finish and alloying element
is present. This is in accordance with the fact that the samples
after heating show a bluish colour, which may be related to the The effect of surface finish and alloying element is already illus-
chromium oxide formed on the sample surfaces. This is the reason trated in Fig. 4 (Section 3.1). However, the comparison for the three
that the emissivity increases with increasing temperature at short alloys and the three surface finishes is better visualized in the graph
wavelength. in Fig. 5, where the emissivity spectra of the nine types of samples
In Fig. 2 the emissivity spectra measured for the sandblasted measured at T ∼ 515 ◦ C are shown. The lowest emissivity is found
samples are shown. As observed, the qualitative behaviour of the for brushed samples and the highest for wire-cut EDM surfaces.
 L. del Campo et al. / Journal of Alloys and Compounds 489 (2010) 482–487 485

Fig. 5. Spectral normal emissivity (ε) of the nine types of samples in Table 2 as a
function of the wavelength () at a temperature of T ∼ 515 ◦ C.

On the other hand, it is remarkable that although the brushed and

sandblasted samples have nearly the same roughness average val-
ues (see Table 2 in Section 2.1), their emissivity is quite different.
This has to be related to the strong unidirectional lay pattern of
the brushed surfaces, that is, the brushed samples have the rough-
ness extended only in one direction (x direction), while sandblasted
samples are rough in both directions (x and y direction), thus the
latter are in fact rougher.
Besides, the emissivity of the brushed samples is nearly the
same, while it is very different for wire-cut EDM surfaces. This
is due to the recast layer on the wire-cut surfaces which may be
different for each sample. Thus, the emissivity measured for the
wire-cut EDM samples is strongly influenced by the recast layer
Fig. 3. Spectral normal emissivity (ε) of wire-cut EDM samples as a function the
wavelength () for several temperatures: (a) Inconel 718 (sample type 3), (b) René
formed during the machining process.
41 (sample type 6), and (c) Haynes 25 (sample type 9).
3.3. Dependence on the emission angle

The angular spectral emissivity has been measured for discrete

angles between the normal and 85◦ using sandblasted samples. This
surface finish has been chosen because of its isotropy, so that only
the polar angle has to be taken into account. In Fig. 6 the angular
emissivity of sandblasted samples measured at T ∼ 455 ◦ C is dis-
played as a function of the emission angle, for several wavelengths.
As seen, the three alloys exhibit a similar angular dependence. At
short wavelengths, the emissivity nearly maintains constant till
60◦ , and then it starts to decrease until zero value is achieved at
90◦ . However, as the wavelength increases, the angular dependence
becomes more “metallic-like”, and above 10 ␮m, the emissivity
increases from normal direction till it achieves a maximum at an
angle close to 90◦ , and then abruptly goes to zero at 90◦ . The long
wavelength behaviour is the expected one for a metallic sample,
as predicted by the electromagnetic theory. Nevertheless, as the
wavelength decreases and due to the surface texture, the angu-
lar dependence of the emissivity of the macroscopic surface differs
from the one of microscopic surface, and the experimental results
are different from the results predicted by the electromagnetic the-
ory.

3.4. Effect of the oxidation

The alloys considered in this paper have a very good corro-

sion resistance. Emissivity measurements during in situ oxidation
processes in air at 700 ◦ C have been performed on brushed sam-
ples, but even after 20 h the emissivity variation due to oxidation
was not so important, mainly at long wavelength. The results
Fig. 4. Spectral normal emissivity (ε) of the nine types of samples in Table 2 plotted
as a function of the temperature at wavelengths (): (a)  = 2.5 ␮m, (b)  = 10 ␮m,
are shown in Fig. 7, where the emissivity spectra measured dur-
and (c)  = 20 ␮m. ing the in situ oxidation process are shown for several oxidation
times.
486 L. del Campo et al. / Journal of Alloys and Compounds 489 (2010) 482–487

Fig. 6. Spectral directional emissivity (ε) of sandblasted samples as a function of the

emission angle for several wavelengths at a temperature of T∼455 ◦ C. (a) Inconel 718
(sample type 2), (b) René 41(sample type 5), and (c) Haynes 25 (sample type 8).
Fig. 7. Spectral normal emissivity (ε) of brushed samples for several oxidation times
measured during an in situ oxidation process in air at 700 ◦ C: (a) Inconel 718 (sample
type 1), (b) René 41(sample type 4), and (c) Haynes 25 (sample type 7).
As seen, when the oxidation time increases, the emissivity rises.
This fact is the expected since the emissivity of the oxides is usually
higher than the emissivity of the metal. The less oxidation resis-
tant alloy seems to be the René 41, while its emissivity is the one
that most has changed after 20 h of oxidation process. This is also
revealed in Figs. 1 and 2 (Section 3.1) where the emissivity increase
with the temperature at short wavelength is more manifest for
René 41 than for Inconel 718 and Haynes 25. Additionally, the max-
imum that appears in the emissivity at short wavelength (Fig. 7b)
is due to radiation interferences in the oxide scale [14]. Even after
20 h of oxidation, the thickness of the oxide film formed on Inconel
718 and Haynes 25 is still too small for the interferences to appear.
This also indicates that the oxide scale is thicker in the René 41 than
in the other two samples.
The oxidation resistance of these alloys is considerably lowered
when exposed to higher temperatures. Thus, a brushed Inconel 718
sample has been oxidized during 30 min in air at 900 ◦ C using a Fig. 8. Spectral normal emissivity (ε) of a brushed Inconel 718 sample pre-oxidized
separate furnace, and afterwards its emissivity has been measured in air at 900 ◦ C during 30 min, as a function of the wavelength () and for several
as a function of the temperature. Fig. 8 shows the obtained nor- temperatures.
mal emissivity spectra. On one hand, the interferencial extremes
are already visible at short wavelength, thus the oxide scale has
grown larger in 30 min at 900 ◦ C than in 20 h at 700 ◦ C. Additionally, 4. Conclusions
the position of the maximum and minimum do not change as the
temperature varies, consequently the sample is not further oxidiz- It has been found that the emissivity is nearly the same for the
ing. On the other hand, the oscillations observed in the emissivity three studied Ni and Co alloys. In general terms, the emissivity
at long wavelength are emission bands due to lattice vibrations decreases with increasing wavelength, and it is nearly independent
of the oxide. It is clear that the phonon frequencies change with of the temperature in the analyzed temperature range. The angular
temperature and the bands broaden due to a rise in the damping emissivity at short wavelengths maintains constant with increas-
parameter of the oscillators. An appropriate modelling of the emis- ing angle till 60◦ and then it becomes zero at 90◦ . In contrast, the
sivity could be performed to obtain the oxide thickness, as well angular emissivity at long wavelengths increases with increasing
as the optical indices and the frequencies of the normal modes emission angle till it achieves a maximum, and then it drops to zero
[14–16]. at 90◦ . Finally, oxidized samples have a higher emissivity, and inter-
 L. del Campo et al. / Journal of Alloys and Compounds 489 (2010) 482–487 487

ferential effects of the radiation due to the oxide scale have been their support through a Ph.D. fellowship. L. del Campo acknowl-
found. edges the University of the Basque Country and the Basque
The spectral directional emissivity values given in this paper Government for their support through Ph.D. and postdoctoral fel-
can be used to measure the temperature of a target by pyrometry. lowships. She also gratefully thanks the “Radiative and Transport
To do so, it is important to make sure that the alloy is not oxi- Properties of Materials” group of the CEMHTI (CNRS) laboratory at
dized, because the emissivity can drastically change with oxidation. Orléans for the useful discussions during her postdoctoral stay.
Finally, it is very important to identify the wavelength where the
radiation sensor of the pyrometer detects, so as to use the emissivity References
at the correct wavelength.
These emissivity values can also be used to simulate thermal [1] R. Siegel, J. Howell, Thermal Radiation Heat Transfer, fourth ed., Taylor & Francis,
Washington, 2002.
radiation heat transfer in an airplane nozzle or a furnace. If nec- [2] G. Tanda, M. Misale, J. Heat Trans. ASME 128 (2006) 302–306.
essary, the total hemispherical emissivity can be calculated by [3] G.A. Greene, C.C. Finfrock, T.F. Irvine Jr., Exp. Therm. Fluid Sci. 22 (2000)
integrating the spectral directional emissivity values given in this 145–153.
[4] G. Pottlacher, H. Hosaeus, B. Wilthan, E. Kaschnitz, A. Seifter, Thermochim. Acta
paper. More than 90% of the energy emitted by a body in the studied 382 (2002) 255–267.
temperatures belongs to the spectral range between 2 and 12 ␮m, [5] S. Alaruri, L. Bianchini, A. Brewington, Opt. Lasers Eng. 30 (1998) 77–91.
thus integration in this spectral range is enough to approximately [6] T. Makino, O. Sotokawa, Y. Iwata, Int. J. Thermophys. 9 (1988) 1121–1130.
[7] M. Kobayashi, M. Otsuki, H. Sakate, F. Sakuma, A. Ono, Int. J. Thermophys. 20
obtain the total emissivity.
(1999) 289–298.
If the aim is to minimize infrared emission for low observability [8] M. Kobayashi, A. Ono, M. Otsuki, H. Sakate, F. Sakuma, Int. J. Thermophys. 20
purposes, it is crucial to control the sample oxidation, and smooth (1999) 299–308.
surfaces are suitable. Thus it is advisable to smooth the metallic [9] W. Bauer, H. Oertel, M. Rink, TEMPMEKO 2001—8th Symposium on Tempera-
ture and Thermal Measurement in Industry and Science, vol. 1, no. 30, Berlin,
surfaces inside the nozzle, and an anti-oxidizing coating is appro- Germany, June 19–21, 2001, p. 306.
priate. Among the three studied alloys, none is better or worse, as [10] W.J. O’Sullivan Jr., W.R. Wade, Theory and Apparatus for Measurement of Emis-
they have nearly the same emissivity. sivity for Radiative Cooling of Hypersonic Aircraft with Data for Inconel and
Inconel X, NACA Technical Note, 4121, 1957.
[11] J.R. Markham, K. Kinsella, R.M. Carangelo, C.R. Brouillette, M.D. Carangelo, P.E.
Acknowledgements Best, P.R. Solomon, Rev. Sci. Instrum. 64 (1993) 2515–2522.
[12] L. del Campo, R.B. Pérez-Sáez, X. Esquisabel, I. Fernández, M.J. Tello, Rev. Sci.
Instrum. 77 (2006) 113111.
This work has been carried out with the financial support [13] R.B. Pérez-Sáez, L. del Campo, M.J. Tello, Int. J. Thermophys. 29 (2008)
of the SAIOTEK program (Project number S-PC08UN07) of the 1141–1155.
Basque Government and the “Universidad-Empresa” program [14] L. del Campo, R.B. Pérez-Sáez, M.J. Tello, Corros. Sci. 50 (2008) 194–199.
[15] D. De Sousa Meneses, J.F. Brun, P. Echegut, P. Simon, Appl. Spectrosc. 58 (2004)
(Project number UE06/01) of the University of the Basque Coun- 969–974.
try in collaboration with “Industria de Turbopropulsores S.A.”. L. [16] T. Iuchi, T. Furukawa, S. Wada, Appl. Opt. 42 (2003) 2317–2326.
González-Fernández acknowledges the Basque Government for