Marc Rger

Plataforma Solar de Almera,

Institute of Technical Thermodynamics,
German Aerospace Center (DLR),
Apartado 39,
E-04200 Tabernas, Spain
Christoph Rickers
Fraunhofer Institute for Thin Films and Surface
Technology IST,
Bienroder Weg 54E,
D-38108 Braunschweig, Germany
Ralf Uhlig
Institute of Technical Thermodynamics,
German Aerospace Center (DLR),
Pfaffenwaldring 38-40,
D-70569 Stuttgart, Germany
Frank Neumann
Christina Polenzky
Fraunhofer Institute for Thin Films and Surface
Technology IST,
Bienroder Weg 54E,
D-38108 Braunschweig, Germany
Infrared-Reective Coating on
Fused Silica for a Solar
High-Temperature Receiver
In concentrating solar power, high-temperature solar receivers can provide heat to highly
efcient cycles for electricity or chemical production. Excessive heating of the fused-
silica window and the resulting recrystallization are major problems of high-temperature
receivers using windows. Excessive window temperatures can be avoided by applying an
infrared-reective solar-transparent coating on the fused-silica window inside. Both glass
temperatures and receiver losses can be reduced. An ideal coating reects part of the
thermal spectrum 2.5 m of the hot absorber 1100C back onto it without re-
ducing solar transmittance. Extensive radiation simulations were done to screen different
lter types. The examined transparent conductive oxides involve a high solar absorp-
tance, inhibiting their use in high-concentration solar systems. Although conventional
dielectric interference lters have a low solar absorption, the reection of solar radia-
tion, which comes from various directions, is too high. It was found that only rugate
lters fulll the requirements for operation under high-ux solar radiation with different
incident angles. A thermodynamic qualication simulation of the rugate coating on a
window of a at-plate receiver showed a reduction of almost 175 K in mean window
temperature and 11% in receiver losses compared with an uncoated window. For the
conguration of a pressurized receiver (REFOS type), the temperature could be reduced
by 65 K with slightly reduced receiver losses. Finally, a 25 m thick rugate lter was
manufactured and optically characterized. The measured spectra tted approximately the
design spectra, except for two absorption peaks, which can be avoided in future deposi-
tions by changing the deposition geometry and by using in situ monitoring. The issue of
this paper is to share the work done on the choice of lter type, lter design, thermody-
namic evaluation, and deposition experiments. DOI: 10.1115/1.3097270
1 Introduction
In central receiver technology, two-axis tracking heliostats con-
centrate solar radiation onto the receiver, which is located on top
of a tower. The high solar concentration factors between 200 and
1000 enable efcient operation of high-temperature receivers,
which normally guarantee high cycle efciencies. An example is
the pressurized volumetric receiver for example REFOS 1,2
and directly irradiated annular pressurized receiver DIAPR
3,4, which integrates the solar energy directly into the topping
Brayton cycle of a combined cycle. These receivers have a trans-
parent window to separate the pressurized air from ambient. Fig-
ure 1 shows details of the REFOS-type receiver.
Open volumetric or tubular receivers may also be equipped
with a window to separate the receiver working medium from
ambient or merely to reduce losses. Frequently, fused silica, hav-
ing a high solar transmittance, low expansion coefcient, and
hence good shock resistance, is used as window material.
One major challenge in high-temperature receiver technology
using fused-silica windows is to avoid recrystallization of the win-
dow. For an enduring safe operation of pressurized volumetric
receivers, a maximal fused-silica temperature of 800C is consid-
ered tolerable.
One approach to avoid window overheating is using active
cooling measures, e.g., air-jet cooling on the window outside 5,6
or gap cooling on the window inside 7. The active measures
involve additional components, such as a blower, air distribution
systems, and nozzles in air-jet cooling, or a second window and an
air distribution ring in gap cooling. Furthermore, active cooling
systems usually affect system efciency.
In contrast to active measures, passive measures reduce the heat
input and hence avoid excessive window heating. Infrared-
reective IRR solar-transparent high-temperature stable coatings
on fused silica belong to the passive measures. Visibly transparent
IRR coatings are widely used in other sectors, e.g., as solar-
control and low-emissivity low-e coatings in architectural glaz-
ing. Also in concentrated solar thermal technology, the use of
coatings on glass is steadily increasing. Antireective coatings on
the receiver for parabolic troughs are state-of-the-art. Research
and development in the eld of high-ux solar thermal application
are on the way, for example, for spectrally selective lters beam
splitters 8 or IRR lters this paper.
2 The Ideal Filter and Operating Principle
Figure 2 shows a section of the geometric receiver congura-
tion and the coating operation principle. The fused-silica window
is arranged in front of the absorber 1100C. The coating is
deposited on the window side facing the absorber surface.
For application on the window of a high-temperature high-ux
solar receiver, the coating must fulll several optical, thermal, and
mechanical requirements. About 98.8% of the solar spectrum air
mass AM 1.5 reected off the heliostat eld has a wavelength
smaller than 2.5 m. To transmit this concentrated radiation
through the receiver window, the wavelength cutoff cannot be
smaller than 2.5 m. However, to reect the maximum fraction
of radiation emitted by the hot absorber back onto it, the cutoff
should be as small as possible. Hence, the optimal wavelength
cutoff is about 2.5 m. Due to the high solar ux densities, it is
Contributed by the Solar Energy Engineering Division of ASME for publication
in the JOURNAL OF SOLAR ENERGY ENGINEERING. Manuscript received August 28, 2007;
nal manuscript received March 17, 2008; published online March 24, 2009. Review
conducted by Akiba Segal. Paper presented at the 2007 ASME Solar Energy Division
and Advanced Energy Systems Division Conference ES2007, Long Beach, CA,
June 2729, 2007.
Journal of Solar Energy Engineering MAY 2009, Vol. 131 / 021004-1 Copyright 2009 by ASME
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms
also essential that the absorptance of the solar spectrum
2.5 m is not increased. Otherwise heat input by solar absorp-
tion increases the temperature prohibitively. To reect the ab-
sorber radiation, reectance for wavelengths 2.5 m should
be maximal.
The coating must be heat-resistant at least up to 800C and
show long-term stability, also under cyclic heating. Additionally, it
must mechanically resist window handling and cleaning and
should not affect the fused-silica structure. For example, it should
not accelerate cristobalite formation.
3 Examined Coating Types
The main tasks in designing optical lters are the selection of
appropriate materials, not only depending on optical properties
but also on mechanical requirements, and thus nding a useful
compromise between technological feasibility and optical perfor-
mance. To realize a coating for a solar receiver window using thin
lm technologies, different approaches are possible.
3.1 Transparent Conductive Oxide Filter. The transparent
conductive oxide lter TCO lter working principle is spectral
excitation. TCO coatings are wide-gap semiconductors, which are
highly doped to get metal-like conductivity due to a high degree
of free charge carriers. At the same time, the wide band gap
avoids excitation of charge carriers into the conductive band,
which ensures transparency within the visible region. Thus, the
functionality can be considered a material property. TCO lms are
widely used, e.g., as transparent electrodes for solar cells or in the
display technology to contact liquid crystal displays or organic
light-emitting diodes.
3.2 Interference Filter. In contrast to TCO lters, the inter-
ference lters solely function by interference effects, which are
created using a sequence of layers made from at least two differ-
ent materials 9. When using dielectric materials without absorp-
tion, lters that are completely transparent for the solar radiation
might be possible.
The design of conventional interference lters is mostly done
by an optimization algorithm, which varies the layer sequence and
layer thicknesses, calculates the theoretical spectra, compares
them to the target, and minimizes the discrepancy between target
and theory until a sufcient correspondence is reached. Conven-
tional interference lters are used in a large variety of applica-
tions, e.g., as antireective coatings, mirrors for laser applications,
or lters band, cut-off, and neutral density in measurement de-
vices or cameras. One major drawback is an effect called angular
shift, which results from the law of refraction, where reectance
and transmittance spectra shift with change of the viewing angle.
While conventional interference lters consist of only two ma-
terials low- and high-refractive, rugate lters lters with cor-
rugated index proles comprise a quasicontinuous mixture of
two materials leading to a continuously modulated index of re-
fraction 10. Figure 3 shows a comparison of exemplary index
proles and Fig. 4 shows the resulting reection spectra of the two
lter types. The rugate interference lter right shows that a
simple apodization by tailoring the index prole corrugation
leads to a suppression of ripples in the reectance and transmit-
tance spectra.
In comparison to conventional lters, rugate lters possess an
increased thermal stability and improve the optical performance
both for normal incidence and oblique/angular-range lters. Dif-
ferent processes for the deposition of rugate lters and a design
approach are presented in Ref. 11.
4 Thermodynamic Evaluation
Table 1 gives an overview of the lters evaluated in this paper.
The uncoated GE 214 substrate is the status quo, the ideal lter
shows the maximal potential, and the TCO and conventional/
rugate interference lters show what can be expected from the
individual technologies.
4.1 Radiation Model. The performance of the individual l-
ter designs on window temperature and receiver loss is evaluated
using a radiation model developed at Deutsches Zentrum fr Luft-
und Raumfahrt DLR. The radiative heat transfer between the
absorber, semitransparent window, and ambient is modeled by ap-
Fig. 1 Pressurized high-temperature receiver REFOS
Fig. 2 Geometric receiver conguration at plate and coat-
ing operation principle
Fig. 3 Index proles of conventional left and rugate lter right in comparison
021004-2 / Vol. 131, MAY 2009 Transactions of the ASME
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms
plying the enclosure method zone method, which was intro-
duced by Hottel and co-worker 12,13. To model the nongray
spectral properties, up to 40 wavelength bands were used. Ab-
sorbed solar radiation is modeled as the source term and is as-
signed a coated and an uncoated window side one-half each. More
details of the model can be found in 6.
Two different receiver geometries were studied with this model.
Zones of large central receivers can be approximated by a at-
plate geometry. This simple at-plate geometry Sec. 5 comprises
of four discretization surfaces one per surface zone: ambient,
window outside, window inside, and ambient. The second re-
ceiver geometry studied was the REFOS receiver Sec. 6. This
REFOS receiver model is discretized in 755 surfaces.
4.2 Coating Evaluation Criteria. From a thermodynamic
point of view, two criteria are relevant for the coating evaluation
on a solar receiver window. The rst criterion is the reduction of
the fused-silica temperature compared with an uncoated substrate.
The temperature of the coated fused silica is essentially deter-
mined by absorption of solar and thermal radiation.
The second criterion for the coating evaluation is the reduction
of receiver losses to ambient. The losses are composed of solar
radiation reected on the window surface and thermal radiation
emitted by window and absorber. For an uncoated window, ther-
mal radiation losses are in the range of two-thirds of the total
losses. Under no circumstances should losses be increased by a
coating. For comparison, the receiver losses listed in the tables are
normalized with the value of the uncoated receiver GE 214 sub-
strate only.
The coatings presented in Sec. 5 are computer designs opti-
mized in that way to get similar spectra as the ideal lter.
5 Evaluation Results of Flat Receivers
The at-plate receiver calculations serve to easily compare the
different lter types without focusing on a special conguration. A
sketch of the at-plate conguration can be seen in Fig. 2. The
boundary conditions of these simulations are listed in Table 2. The
assumption of diffusely incident solar radiation is conservative,
since reectance and absorptance usually increase with increasing
incidence angle.
5.1 The Ideal Filter. Figure 5, rst line, shows the GE 214
fused-silica spectra under diffuse radiation, and the solar and
1100C-blackbody spectra. The GE 214 spectra are based on
measurements 14. Figure 6, rst and second bars, visualizes the
solar weighted absorptance, reectance, and transmittance of the
uncoated 5-mm thick GE 214 fused-silica substrate and the sub-
strate coated with an ideal lter, respectively diffuse radiation.
The solar spectrum air mass 1.5 includes spectral absorption in
the heliostat surface facets. Solar absorptance and reectance are
Fig. 4 Reection spectra of conventional left and rugate lter right in comparison.
The apodization leads to a suppression of ripples.
Table 1 Substrate and lter details
Abbreviation Substrate Coating
GE 214 substrate GE 214 None, substrate: 5 mm fused silica GE 214, re polished
Ideal lter GE 214 Ideal edge lter without changing the substrate for 2, 5 m,
99.5% reectance, and 0.5% absorption for 2, 5 m
TCO lter GE 214 Two TCO layers of ZnO:Al with dielectric cover/packaging layer
Conventional interference lter GE 214 SiO
2
and Ta
2
O
5
70 layers, 6 m
Rugate interference lter GE 214 SiO
2
and Ta
2
O
5
1000 virtual layers, 40 m
Table 2 Boundary conditions for at-plate receiver
Boundary condition Value
Geometry Flat-plate absorber and 5-mm thick window innite extension,
four discretization surfaces
Substrate and coating Spectra from complex refractive indices
Incidence angles Solar: diffuse incidence
Thermal: diffuse incidence
Solar ux density 650 kW/ m
2
homogenous
Absorber temperature 1100C homogenous
Ambient temperature 25C
Convection No convective heat ows on window
Journal of Solar Energy Engineering MAY 2009, Vol. 131 / 021004-3
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms
slightly increased in the ideal lter because a small part of the
spectrum lies above 2.5 m.
Figure 7, visualizes the same for a 1100C-blackbody weight-
ing. According to Fig. 7, a 1100C-blackbody spectrum is trans-
mitted to 54.9%, reected at 12.5%, and absorbed at 32.6% by the
uncoated fused-silica substrate. An ideal lter with a cutoff wave-
length of =2.5 m almost does not change solar transmittance,
but reduces the blackbody-weighted transmittance to 36.5%, in-
creases reectance up to 63% and thus reduces absorptance to less
than 1%. Hence, an ideal lter minimizes both glass temperatures
and receiver losses. The lter reectance in the range of
2.53.5 m reduces thermal transmission losses, while the re-
ectance for wavelengths higher than 3.5 m reduces absorption
of thermal radiation.
For the at-plate conguration, an ideal lter reduces the win-
dow temperatures by 370 K or 390 K mean and maximum tem-
perature, respectively. The mean window temperature is averaged
over the window inside and outside. The maximum temperature is
located on the window inside. The receiver losses are reduced by
28%. Table 3 lists more details.
5.2 Transparent Conductive Oxide Filter. Several exem-
plary computer designs were done using material data from
ZnO:Al TCO lms developed for a high transparency of up to 780
nm and improved conductivity. The material data were acquired
using multi-angle spectral ellipsometry within the spectral region
from 300 nm to 850 nm. The outcome of this method is a disper-
sion model with parameters tted to the measured properties. This
Fig. 5 Absorptance, reectance, and transmittance spectra of the uncoated GE 214 substrate, TCO
lter system, and conventional and rugate interference lters for a at receiver under diffuse radia-
tion. Glass side, left and coating side, right. Filter details, see Table 1.
021004-4 / Vol. 131, MAY 2009 Transactions of the ASME
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms
physical model DrudeLorentz oscillators is used to approxi-
mate the material properties of ZnO:Al for wavelengths above
850 nm. For ZnO:Al TCO material with near metal properties at
energies below the plasma frequency, this should be a reasonable
approximation.
A twin-layer ZnO:Al lm embedded in three dielectric layers is
chosen for thermodynamic evaluation. The spectra under diffuse
radiation is shown in Fig. 5, line 2. The double layer arrangement
was used to get better suppression of transmittance within the IR
part of the spectrum, while keeping high transparency within the
visible spectrum. Furthermore, due to the high operating tempera-
tures, a blocking layer to avoid degradation in the lms by oxida-
tion is needed.
However, the absorptance spectrum shows a relatively high ab-
sorption even within the solar range, where 18.2% of the concen-
trated solar radiation is absorbed in the coated window see Fig.
6. This leads to a high heat input, increasing mean window tem-
peratures to 1068C, i.e., approximately 250 K above the tem-
perature of an uncoated window. Hence, this coating is unsuitable
for the solar high-ux application. The reason for the high solar
absorptance might be the high plasma frequency considering the
targeted application for the evaluated layer. A material develop-
ment using different doping grades or other TCO materials might
lead to a more suitable optical performance. However, a signi-
cant additional effort to cover the TCO materials and to make
them thermally stable would be needed. Furthermore, it is ques-
tionable whether a suitable packaging is feasible at all. Thus, the
time consuming and difcult task of material development was
not further followed.
5.3 Conventional Interference Filter. Interference lters of
up to 100 layers with a thickness of about 10 m were designed
using commercially available thin lm software tools. The absorp-
tance of the dielectric materials SiO
2
and Ta
2
O
5
is as low as
expected, leading to only 1.5% absorption of the solar radiation
see Fig. 6. The strong increase around 3.5 m is mainly due to
the substrate absorption. However, even by use of more layers and
larger lter thicknesses, it was not possible to reduce the ripples in
the reectance in the range of the solar spectrum Although, com-
pared with the uncoated window, the temperature was reduced
slightly by about 35 K, 28.2% of the incident solar radiation is
reected off the window. This leads to too high receiver losses,
which do not compensate for the small temperature reduction. For
comparison, you may see Figs. 57 and Table 3.
5.4 Rugate Interference Filter. The 40-m thick IRR rugate
lter shows only low solar absorptance 1.4%, Fig. 6. The reec-
tance in the solar range is also signicantly reduced by half com-
pared with the conventional approach 14.8%, Fig. 6. Further-
more, the angular dependency is reduced as well. The upper
reectance band range is 6.5 m, the band is more pronounced as
in the conventional approach, and the lter also works for larger
Fig. 6 Solar spectrum weighted absorptance, reectance, and transmit-
tance; glass side
Fig. 7 1100C-blackbody spectrum weighted absorptance, reectance,
and transmittance; coating side
Journal of Solar Energy Engineering MAY 2009, Vol. 131 / 021004-5
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms
angles of incidence. This results in a reectance of the
1100C-blackbody spectrum of 47.2% Fig. 7. The mean/
maximal window temperature can be reduced by 174 K/184 K,
respectively, compared with the uncoated substrate. The receiver
losses decrease by 11% see Table 3.
Of course, this lter cannot be directly compared with the con-
ventional approach in Sec. 5.3 due to the discrepancy within the
technological effort 5 m against 40 m. However, even
thicker conventional lters still reveal the discussed problems.
Additionally, rugate lters should provide better temperature sta-
bility. For the high-ux high-temperature solar receiver window,
rugate interference lters seem to perform best.
6 Evaluation Results of REFOS-Type Receiver
In this section, the simulations for a real high-temperature re-
ceiver geometry are presented, the REFOS receiver. Figure 8
shows its geometry and heat ows. Table 4 lists the boundary
conditions used for the simulations. Due to the cavity shape and
by using incidence angles from ray-tracing, the solar incidence
angles change signicantly compared with that in Sec. 5 assump-
tion of a diffusely radiated at plate, and hence reectance should
be reduced. Radiation heat exchange between absorber, window,
and ambient is also inuenced by the cavity geometry. Heat ex-
change by thermal radiation is assumed to be diffuse. In addition,
convective heat ows due to free convection on the window out-
side and a forced convection due to the hot absorber ow on the
window inside are considered. The thermodynamic radiation
model is discretized into 755 surface elements.
Simulations conrmed that the amount of solar radiation re-
ected off the concave REFOS window is only 2% of the incident
power compared with 13.6% for the at-plate conguration un-
der diffuse radiation, both uncoated windows. The temperature of
the window top reaches 988C due to its exposed position to the
absorber. The receiver losses are 11% of the incident power on the
receiver module secondary outlet plane.
Simulated performance results of the different lter types used
in the REFOS receiver geometry are presented in Table 5, except
for TCO lters, which were already considered as unsuitable in
Sec. 5. The ideal lter reduces maximum window temperatures by
almost 210 K compared with the uncoated window. Losses were
reduced by 25%. The reduction potentials are smaller compared
with the at-plate case of Sec. 5 because of the altered geometry.
The conventional interference lter doubles receiver losses and
reduces maximum temperatures only slightly and hence is, as ex-
pected, unacceptable for this application. The rugate interference
lter reduces maximum temperatures by 78 K and receiver losses
by 4% compared with the uncoated window.
As can be seen in Table 5, the goal of a maximum allowable
fused-silica window temperature of 800C has not been reached
totally. Assuming a lower mean absorber temperature of 1000C,
simulations result in mean window temperatures of 797C and
maximum internal temperatures of 860C. Hence, the maximum
temperature level would be exceeded only by the hottest spot of
the window inside. Mean temperatures would be tolerable with
the designed rugate lter.
To summarize, the potential reduction in window temperature
and receiver loss is higher in a at-plate geometry than in the
REFOS conguration. However, an application of a Rugate lter
on a REFOS receiver would still reduce temperatures and losses.
A rst experimental lter was deposited within this work to test
the technical feasibility of thick rugate lters. SiO
2
and Ta
2
O
5
were deposited simultaneously by rf magnetron cosputtering. The
complete deposition process of a 25-m thick rugate lter lasted
for 100 h. The agreement between expected and measured spectra
is good. The observed problems of deposition rate drift and coat-
Table 3 Evaluation results of at-plate receiver
Abbreviation
Window temp.
C
Normalized
receiver loss Mean Max
GE 214 substrate 821 850 1.00
Ideal lter 450 459 0.72
TCO lter 1068 1117 0.92
Conventional interference lter 787 815 1.30
Rugate interference lter 647 666 0.89
Fig. 8 Thermodynamic model of REFOS receiver
Table 4 Boundary conditions for REFOS receiver
Boundary condition Value
Geometry 3D REFOS receiver geometry, 5-mm thick window see Fig. 8,
755 discretization surfaces
Substrate and coating Spectra from complex refractive indices
Incidence angles Solar: traced rays depending on heliostat and secondary conguration
Thermal: diffuse incidence
Solar radiation and ux density HELIOS simulation: March 21, 12:00 h UTC, 72 heliostats,
solar ux nonhomogenously distributed mean: 570 kW/ m
2

On secondary outlet plane: 464 kW

Absorber temperature 1100C except for inlet absorber ring: 900C
Ambient temperature 25C
Convection Window outside: free convection
Window inside: forced convection
Table 5 Evaluation results of REFOS-type receiver
Abbreviation
Window temp.
C
Normalized
receiver loss Mean Max
GE 214 substrate 908 988 1.00
Ideal lter 745 779 0.75
Conventional interference lter 884 959 2.04
Rugate interference lter 843 910 0.96
021004-6 / Vol. 131, MAY 2009 Transactions of the ASME
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms
ing defects due to parasitic particles falling onto the substrate
surface can be avoided by using in situ monitoring strategies and
other sputter arrangements e.g., a sputter-up geometry.
7 Conclusion and Outlook
Different types of IRR lters on a fused-silica window of a
high-temperature high-ux solar receiver have been studied nu-
merically. The examined TCO lters ZnO:Al are evaluated as
unsuitable because of their high absorptance in the solar spectrum
leading to high window temperatures.
Conventional dielectric interference lters increase solar reec-
tance signicantly for the sun rays coming in from different
angles. The resulting receiver losses are not acceptable, and tem-
perature reduction is only small.
The examined rugate lter designs showed both low solar ab-
sorptance and low solar reectance, also under diffuse radiation.
Simulations with the at-plate receiver geometry showed that the
mean/maximal temperature could be reduced by 174 K/184 K and
the receiver losses by 11%, compared with the uncoated substrate
1100C absorber temperature. The simulations for the REFOS
receiver geometry resulted in lower window temperature reduc-
tions due to the cavity shape of the receiver. However, maximum
temperature still could be decreased by 78 K, and the mean tem-
perature by 65 K 1100C absorber temperature. Up to mean
absorber temperatures of 1000C, the resulting rugate-coated
window temperature would be tolerable without any active cool-
ing systems. The receiver losses were reduced slightly.
A rst 25-m thick lter was deposited by rf magnetron sput-
tering to gain experience in the deposition of thick rugate lters.
For future depositions, an in situ monitoring strategy and a sputter
arrangement preventing parasitic particles from falling down on
the substrate are recommended.
Acknowledgment
Financial support from the German Federal Ministry for the
Environment, Nature Conservation and Nuclear Safety CoMinit
Contract No. 16UM0017 is gratefully acknowledged. Thanks to
M. Lappschies Laserzentrum Hannover for the rugate lter de-
signs.
Nomenclature
A

spectral absorptance
d thickness nm
n refractive index
R

spectral reectance
T

spectral transmittance
wavelength m
Subscripts and Superscripts
S coating side
Abbreviations
REFOS receiver for solar fossil power plants
References
1 Buck, R., Bruning, T., Denk, T., Pfnder, M., Schwarzbzl, P., and Tllez, F.,
2002, Solar-Hybrid Gas Turbine-Based Power Tower Systems REFOS,
ASME J. Sol. Energy Eng., 1241, pp. 29.
2 Heller, P., Pfnder, M., Denk, T., Tllez, F., Valverde, A., Fernandez, J., and
Ring, A., 2006, Test and Evaluation of a Solar Powered Gas Turbine Sys-
tem, Sol. Energy, 8010, pp. 12251230.
3 Karni, J., Kribus, A., Doron, P., Rubin, R., Fiterman, A., and Sagie, D., 1997,
The DIAPR: A High-Pressure, High-Temperature Solar Receiver, ASME J.
Sol. Energy Eng., 119, pp. 7478.
4 Kribus, A., Doron, P., Rubin, R., Reuven, R., Taragan, E., Duchan, S., and
Karni, J., 2001, Performance of the Directly-Irradiated Annular Pressurized
Receiver DIAPR Operating at 20 Bar and 1200C, ASME J. Sol. Energy
Eng., 123, pp. 1017.
5 Rger, M., Buck, R., and Mller-Steinhagen, H., 2005, Numerical and Ex-
perimental Investigation of a Multiple Air Jet Cooling System for Application
in a Solar Thermal Receiver, ASME J. Heat Transfer, 1278, pp. 863876.
6 Rger, M., Pfnder, M., and Buck, R., 2006, Multiple Air-Jet Window Cool-
ing for High-Temperature Pressurized Volumetric Receivers: Testing, Evalua-
tion, and Modeling, ASME J. Sol. Energy Eng., 1283, pp. 265274.
7 Uhlig, R., and Rger, M., 2004, Development of a Window Cooling for High
Temperature Solar Receivers, 22nd CAD-FEM Users Meeting 2004, Inter-
national Congress on FEM Technology With ANSYS CFX & ICEM CFD
Conference, Dresden, Germany, Nov. 1012.
8 Imenes, A. G., Buie, D., Mills, D. R., Schramek, P., and Bosi, S. G., 2006, A
New Strategy for Improved Spectral Performance in Solar Power Plants, Sol.
Energy, 80, pp. 12631269.
9 Furman S. H., and Tikhonravov, A. V., 1992, Basics of Optics of Multilayer
Systems, Editions Frontiers, Gif-sur-Yvette, France.
10 Lange, S., Bartzsch, H., Frach, P., and Goedicke, K., 2006, Pulse Magnetron
Sputtering in a Reactive Gas Mixture of Variable Composition to Manufacture
Multilayer and Gradient Optical Coatings, Thin Solid Films, 502, pp. 2933.
11 Janicki, V., Gbler, D., Wilbrandt, S., Leitel, R., Stenzel, O., Kaiser, N., Lapp-
schies, M., Grtz, B., Ristau, D., Rickers, C., and Verghl, M., 2006, Depo-
sition and Spectral Performance of an Inhomogeneous Broadband Wide-
Angular Antireective Coating, Appl. Opt., 45, pp. 78517857.
12 Hottel, H. C., 1954, Radiant-Heat Transmission, Heat Transmission, 3rd ed.,
W. H. McAdams, ed., McGraw-Hill, New York, pp. 55125.
13 Hottel, H. C., and Sarom, A. F., 1967, Radiative Transfer, McGraw-Hill,
New York.
14 Manara, J., Arduini-Schuster, M., and Amthor, S., 2006, Determination of the
Complex Refractive Index of Glass, Report No. ZAE 2-1004-05.
Journal of Solar Energy Engineering MAY 2009, Vol. 131 / 021004-7
Downloaded From: http://solarenergyengineering.asmedigitalcollection.asme.org/ on 04/21/2014 Terms of Use: http://asme.org/terms