See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312192177

DTU-ESA millimeter-wave validation standard antenna – requirements and

design

Article · January 2014

CITATIONS READS

6 125

8 authors, including:

Kim Branner Rolf Jorgensen

Technical University of Denmark Ticra
132 PUBLICATIONS 2,155 CITATIONS 30 PUBLICATIONS 204 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Kim Branner on 10 January 2017.

The user has requested enhancement of the downloaded file.

Downloaded from orbit.dtu.dk on: Jan 10, 2017

DTU-ESA millimeter-wave validation standard antenna – requirements and design

Pivnenko, Sergey; Kim, Oleksiy S.; Breinbjerg, Olav; Branner, Kim; Markussen, Christen Malte;
Jørgensen , R.; Larsen, Niels Vesterdal; Paquay, M
Published in:
Proceedings of the 36th Annual Symposium of the Antenna Measurement Techniques Association (AMTA)

Publication date:
2014

Link to publication

Citation (APA):
Pivnenko, S., Kim, O. S., Breinbjerg, O., Branner, K., Markussen, C. M., Jørgensen , R., ... Paquay, M. (2014).
DTU-ESA millimeter-wave validation standard antenna – requirements and design. In Proceedings of the 36th
Annual Symposium of the Antenna Measurement Techniques Association (AMTA) . IEEE.

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal ?

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
 DTU-ESA Millimeter-Wave Validation Standard
Antenna – Requirements and Design

S. Pivnenko, O.S. Kim, O. Breinbjerg K. Branner, C.M. Markussen

Department of Electrical Engineering Department of Wind Energy
Technical University of Denmark Technical University of Denmark
2800 Kgs. Lyngby, Denmark 4000 Roskilde, Denmark
sp@elektro.dtu.dk

R. Jørgensen, N. Vesterdal M. Paquay

TICRA ESA/ESTEC
1201 Copenhagen, Denmark 2201, Noordwijk, the Netherlands

Abstract — Inter-comparisons and validations of antenna campaign is achieved when a dedicated Validation Standard
measurement ranges are useful tools allowing the detection of (VAST) antenna specifically designed for this purpose is
various problems in the measurement procedures, thus leading to available. The widespread use of the known VAST-12 antenna
improvements of the measurement accuracy and facilitating (working at 12 GHz), developed by the Technical University of
better understanding of the measurement techniques. The Denmark (DTU) and operated by the DTU-ESA Facility since
maximum value from a validation campaign is achieved when a 1992, demonstrates the long-term value of the dedicated VAST
dedicated Validation Standard (VAST) antenna specifically antennas [5]. The VAST-12 antenna has been used extensively
designed for this purpose is available. The driving requirements for measurement facility comparison and validation over many
to VAST antennas are their mechanical stability with respect to
years; in particular within the EU Network of Excellence ACE
any orientation of the antenna in the gravity field and thermal
– Antenna Centre of Excellence which led also to important
stability over a given operational temperature range. In addition,
VAST antennas must possess electrical characteristics that are recommendations for future VAST antennas and comparison
typical for satellite antennas and challenging to measure. campaigns [5].

A multi-band millimeter-wave VAST (mm-VAST) antenna for The driving requirements of VAST antennas are their
the K/Ka-bands and Q/V bands is currently under development mechanical stability with respect to any orientation of the
in collaboration between the Technical University of Denmark antenna in the gravity field and thermal stability over a given
(DTU) and TICRA under contract from the European Space operational temperature range. The mechanical design shall
Agency. In this paper, the electrical and mechanical ensure extremely stable electrical characteristics with variations
requirements of the DTU-ESA mm-VAST antenna are discussed typically an order of magnitude smaller than the measurement
and presented. Potential antenna types fulfilling the electrical uncertainty. At the same time, it must withstand high g-loads
requirements are briefly reviewed and the baseline design is under frequent transportations and it shall also support
described. The emphasis is given to definition of the requirements convenient handling of the VAST antenna (practical electrical
for the mechanical and thermal stability of the antenna, which and mechanical interfaces, low mass, attachment points for
satisfy the stringent stability requirement for the mm-VAST lifting, etc.). Also, the VAST antenna must allow for precise
electrical characteristics. definition of at least one coordinate system to which all pattern
measurements must be referred.
I. INTRODUCTION
Inter-comparison and validation of antenna measurement Today, there is a well identified need for increased
ranges either with readily available antennas or with dedicated operational frequencies to get access to large bandwidth for
VAST antennas have been carried out for at least three decades broadband communication. Upcoming satellite communication
[1-5]. These activities have proven to be very useful as these services utilize up/down link at K/Ka-bands, while the use of
allow finding and help correcting major and minor problems in Q/V bands is contemplated for the feeder links in the coming
the measurement procedures, thus leading to an improvement years. In response to this need, a multi-band millimeter-wave
of the measurement accuracy of the involved ranges and VAST (mm-VAST) antenna covering K/Ka/Q/V-bands is
facilitating better understanding of the measurement currently under development in collaboration between DTU
techniques. and TICRA under contract from the European Space Agency.

The early experience gained from the campaigns In this paper, the electrical and mechanical requirements of
demonstrated that readily available antennas can be used for the DTU-ESA mm-VAST antenna are discussed and presented.
this purpose, but they only partially meet the requirements of a Potential antenna types fulfilling the electrical requirements are
validation campaign. The maximum value from a validation briefly reviewed and the baseline design is described. The
emphasis is given to definition of the requirements for the The cross-polar pattern shall have a null in the
mechanical and thermal stability of the antenna, which satisfy main beam; quantified as a level of some 50 dB
below the co-polar peak.
the stringent stability requirement for the mm-VAST electrical
characteristics. Polarization Reconfigurable between linear and circular
polarization for all operational frequencies.
II. REQUIREMENTS TO MM-VAST ANTENNA Operational One frequency within each of
Band 1: 17.5-20.2 GHz
In order to increase the outcome from facility comparisons, frequencies Band 2: 27.5-31.0 GHz
the VAST antenna electrical characteristics should possess a Band 3: 37.5-40.5 GHz
Band 4: 47.2-50.2 GHz
series of features combining both representative characteristics These bands comply with the Primary allocations
of typical satellite antennas and complex details challenging the for Fixed Satellite Service.
measurement. Many typical errors in antenna measurement Gain Within 30-35 dBi for bands 1 and 2 and within
ranges can be seen from the location and shape of the pattern 33-38 dBi for bands 3 and 4.
peak, the shape and level of the near and far sidelobes, the Input impedance The reflection coefficient shall be between
cross-polar pattern, the location and level of the nulls. It is also -10 dB..-20 dB for the operational frequencies.
preferred that the pattern is asymmetric, with its peak slightly Feed No feed horn mounting/dismounting is allowed
offset from the axes of the coordinate system allowing arrangement
when changing the frequency band.
detection of problems in its definition. Waveguide components are allowed to be
remounted for changing the frequency band
It was observed in the facility comparison campaign carried and/or polarization.
out with the VAST-12 antenna [5] that scattering from the Electrical The electrical interface shall be standard
rectangular waveguides:
antenna positioner may have noticeable effect when its interface • WR42 for band 1
illumination is significant. Thus the VAST antenna radiation • WR28 for band 2
towards its support structure should be minimized by proper • WR28 or WR22 for band 3
design or by proper location of the mounting flange. • WR22 or WR19 for band 4
Stiffness and The antenna shall be rigid enough to withstand
Another very important requirement is the high stability of the amount of handling and transportation to
electrical characteristics ensuring that the differences in the robustness which it will be subjected.
measurement results from different measurement ranges and Mechanical and The antenna shall be stiff enough such that any
techniques are not coming from the deformations of the deformities of the antenna during test, assuming
thermal stability that the antenna is to be rotated, introduce an
antenna itself. Stability of the electrical characteristics of the
error not larger than 0.003 dB (1σ) in the peak
antenna is directly dependent on its mechanical stability, e.g. directivity.
structural deformations due to gravity changes when the
Temperature Operational temperature range from 15ºC to 25ºC
antenna is rotated, and its thermal stability, e.g. antenna Survivability temperature range -10ºC to +50ºC
and humidity
thermal expansion due to changed temperature. To this end, Long time relative humidity limits 40% to70%
ranges
variations of the antenna characteristics due to the above
factors an order of magnitude smaller than the expected Weight Less than 50 kg.
measurement uncertainty are required. In the state-of-the-art Dimensions No specific requirement.
measurement facilities, an achievable measurement uncertainty
for an antenna with some 30-40 dBi directivity is typically Coordinate Optical CS defined by a mirror cube.
about 0.03 dB (1σ) referenced to the peak directivity [5, 6]. system definition Mechanical CS defined by the mounting flange
and spirit level.
Thus the value characterizing the electrical stability of the Hardware Use of components-of-the-shelf is preferred.
VAST antenna is taken to be 0.003 dB (1σ). development Compatible with class 100.000 .

The main electrical and mechanical requirements to the Handling and Necessary lifting and handling tools. Shock level
mm-VAST antenna are summarized in Table I. transportation detection sensors included. Dedicated transport
container.
TABLE I. ELECTRICAL AND MECHANICAL REQUIREMENTS. Mech. interface Compatible with the VAST-12 antenna.

Parameter Description
Co-polar pattern Near sidelobes (1st-3rd) shall be present in the III. ANTENNA TYPES SUITABLE FOR MM-VAST
range 18-25 dB below the pattern peak. Deep
nulls are desirable. Due to the main requirements to the mm-VAST antenna −
The radiation towards a support structure (to be such as high mechanical and thermal stability, durability,
specified) shall not introduce uncertainty larger minimum research and development (R&D), maximum use of
than 0.01 dB (1σ).
An asymmetry visible in the first few sidelobes components-of-the-shelf (COTS) − certain antenna types can
shall be present at least in one plane of the be immediately ruled out:
radiation pattern. The beamwidths in the
orthogonal planes shall differ. Flat-top or split • Waveguide/waveguide-slot arrays – involves large R&D,
main beam in one of the planes. COTS can hardly be used, challenging to make multi-band;
Cross-polar The cross-polarization level in the main beam
shall be lower than 20 dB below the co-polar • Printed phased arrays – in addition to the problems above,
pattern level within the −1 dB contour at a selected a PCB substrate can be a problem with respect to the
operational frequency in each band. thermal stability as well as durability;
• Printed reflectarrays – same problems as above; by the increased focal distance to diameter (f/D) ratio. During
the initial design phase of the VAST-12, this configuration
• Wideband horns cannot be used due to their low
went through a thorough comparison study (together with a
directivity. dual reflector, prime focus configuration) and was selected as
Thus, a reflector antenna presents a natural choice for the the most appropriate one. Subsequent development, tests, and
mm-VAST. The experience gained from the development and operation of the VAST-12 have confirmed the validity of this
operation of the VAST-12 confirms that a reflector antenna can choice.
be made mechanically and thermally stable, durable for Therefore, a single reflector, offset configuration, has
decades, and at the same time the R&D work is reduced to formed the baseline for the development of the mm-VAST.
minimum, as the reflector and the feed support structure are the
only parts, which need to be designed.
IV. FEED CONFIGURATION
A. Single reflector, prime focus As the mm-VAST antenna has to operate at four widely
This configuration exhibits a good cross-polar performance. separated frequencies, within the Ku-, Ka- Q-, and V-bands,
However, significant scattering from the feed(s) and the the first quite natural choice is to use a dedicated narrowband
massive feed support structure can increase the far-out feed for each frequency.
sidelobes to an unacceptable level. Attempts to minimize these
Another option is to use a single wide-band feed covering
effects will compromise the mechanical stability of the support
all four bands of interest. However, known wideband antennas
structure. Moreover, additional uncertainties are introduced in
that can be used as the feed will require a custom designed
the measurements due to feed cables arbitrarily located in the
transition to a standard waveguide for each of the four bands.
high field of the main reflector.
This is associated with a lengthy development as well as costly
B. Dual reflector, prime focus manufacturing and testing and thus not considered suitable
It potentially exhibits excellent cross-polar performance as within this project.
well as very low spill-over losses and back radiation. The The possibility of using only two conical horns, with each
support structure carrying only sub-reflector can be made one working in two neighboring bands was also considered, but
simpler and lighter than for a feed, and thus the pattern due to large difference of the beamwidth and possible
disturbing effects due to the scattering are less significant. Feed excitation of the higher-order waveguide modes this
ports and cables are hidden behind the main reflector, thus configuration was discarded.
reducing the measurement uncertainty. On the other hand, this
configuration requires highly directive feed(s), and thus large Yet another option is to use two specially designed
feed aperture(s) and larger offsets from the focal point, if a wideband corrugated horns, each one covering two neighbor
multiple-feed set-up is chosen. Compared to a single reflector bands, with approximately the same beamwidth in each band.
configuration, a dual reflector system is inherently more costly The designs of such horns with 1:2 relative bandwidth are
in design and manufacturing. Another disadvantage is the feed available for 20-30 GHz [7, 8]. Detailed analysis of this
return loss due to the feed sub-reflector interaction. This is configuration has shown only little improvement over the four
especially a problem for small main reflector diameters narrowband feed and thus it was ruled out.
(expected diameter in this case is 20λ-30λ). In conclusion, it was found that the best solution is to use 4
C. Dual reflector, offset conical horns each one working in one of the 4 frequency
bands.
The feed sub-reflector scattering problems are reduced
significantly as no reflected (only diffracted) rays will hit the V. ELECTRICAL BASELINE DESIGN
feed and sub-reflector. However, the offset geometry will A series of extensive simulations for an offset reflector was
create high cross polarization on either side of the symmetry carried out using GRASP software [9] with varying parameters
plane. This cross-polarization can be reduced by applying cross such as reflector diameter, focal distance, feed offset, reflector
polarization compensated geometries such as Gregorian, front- surface shape, rim shape, and feed configuration. The obtained
fed Cassegrain and side-fed Cassegrain. On the other hand, the results were summarized in a compliance matrix versus the
first configuration will give rise to higher side lobes in the ±20° electrical requirements described in Table I.
range, due to the feed/sub-reflector radiation directly to the far
field, while in the latter two configurations, the feed direct For the chosen preliminary reflector parameters, all
radiation will increase the sidelobes in the back and side considered feed configurations showed rather similar
region, respectively. Furthermore, the manufacturing cost of performance potentially meeting the requirements. However,
the front-fed or side-fed Cassegrain antennas is more than when using the wideband horns or conical horns in the
twice of that of a single reflector system, as the main and sub- neighbor bands, excitation of the higher-order modes in the
reflectors are nearly of the same size, which in addition require higher-frequency band may potentially give rise to problems
a bulky frame to ensure mechanical stability. with repeatability of the results after changing from linear (LP)
to circular polarization (CP) and from band to band. Thus the
D. Single reflector, offset configuration with four narrow-band feeds arranged along the
This configuration does not suffer from the disadvantages symmetry plane of a dual-focus reflector with a rectangular
of the single-reflector, prime focus configuration. The cross- aperture was finally selected, see Fig 1. The parameters of the
polar performance is worse, which however can be mitigated optimized configuration are summarized in Table II.
 Figure 1. Optimized configuration of the mm-VAST
antenna: CGL – global coordinate system, CFE – feed cluster
coordinate system.

TABLE II. OPTIMIZED REFLECTOR PARAMETERS.

Further analysis has shown that using Potter horns allows

significant improvement of the cross-polar performance at CP.
Thus the final feed configuration was selected as being a
cluster of four Potter horns with the apertures placed in the
same plane, while the horn axes are parallel and located in the
antenna symmetry plane. The order of the 4 feeds is chosen to
minimize the electrical offset from the focal point in the offset
plane; thus, the V-band horn is located at the focal point, with
the Q-band and Ka-band horns on each side, and the K-band
horn next to the Q-band horn.
The radiation patterns around the main beam and first few
sidelobes for the LP and CP and for all four bands are shown in
Fig. 2 and Fig. 3.
It is seen from Figs. 2-3 that the electrical requirements in
Table I are satisfied in this baseline design. The only comment
is that for LP the level of cross-polarization is just below the
goal of −20 dB in the bands 2 and 4. It is expected that the feed
support structure may result in degradation of this value, and
some further fine tuning of the reflector parameters is planned
taking into account the effects of the feed support frame. On
the other hand, it is noted that for the strongly elliptic shape of
the main beam it is rather difficult to ensure this cross-polar
level within the -1 dB contour, even though the on-axis cross-
polar level has a deep null.
Figure 2. Simulated mm-VAST patterns for the LP.
 VI. REQUIREMENTS TO THE MECHANICAL AND THERMAL
STABILITY
The requirement to the electrical stability of the mm-VAST
antenna is set to be 0.003 dB (1σ) referenced to the peak
directivity. From this value, the requirements to the mechanical
stability of the feed support structure, i.e. maximum acceptable
displacement of the feed with respect to its nominal position
has to be determined.
Feed displacement orthogonal to its symmetry axis causes
mainly a reflector main beam shift proportional to the feed
displacement angle. A series of simulations were carried out for
the measured patterns of the VAST-12 antenna having about
30 dBi directivity and a reflect-array antenna [8] having about
40 dBi directivity. The far-field patterns of these antennas were
calculated in a coordinate system shifted by 0.01º and 0.003º
for the VAST-12, and by 0.005º and 0.002º for the reflect-
array, separately in a symmetry plane and in an offset plane. A
complex difference between the far fields obtained with the
nominal and displaced feed was then calculated. This
difference in the logarithmic scale represents the equivalent
error (difference) signal (EES) between the fields from the
nominal feed position and the displaced feed. The EES level
corresponding to the 0.003 dB (1σ) pattern uncertainty is
−60 dB. The VAST-12 radiation pattern in the offset plane with
the EES level corresponding to the 0.003º pattern displacement
is shown in Fig. 4.

Figure 4. VAST-12 offset plane, 0.003º displacement.

It is seen from Fig. 4 that for the VAST-12 antenna for the
feed displacement of 0.003º the EES level reaches the −60 dB
limit, while for the 40 dBi antenna (not shown here), this limit
is achieved at about 0.0017º feed displacement. The calculated
maximum angular feed deviations can be transferred to
maximum feed translation by calculating tangent of the found
angle times the focal distance (assuming the deformation
happens in one point where the feed support is attached to the
reflector). For the focal distance of 167 mm, it is:
167×tan(0.0017º) = 0.005 mm = 5 µm.
A similar investigation for the feed displacement carried
out with GRASP have shown very similar maximum feed
displacement values orthogonal to the feed axis, but much
smaller value, 1 µm, for the feed displacement along the feed
axis for the same EES level at −60 dB. Such tight tolerance is
very difficult to achieve from the mechanical point of view and
Figure 3. Simulated mm-VAST patterns for the CP. thus another series of simulations was carried out to clarify the
effect of the longitudinal feed displacement on the far-field
pattern obtained from a typical near-field antenna allowed error level. The results for the 5º rotated main beam
measurement. (not shown here) have shown even higher sensitivity for the θ-
scanning scheme: only 0.25º total phase shift is allowed.
Taking the far-field pattern of the VAST-12 antenna
obtained from a spherical near-field measurement, the At 50 GHz with 6 mm wavelength the 0.25º phase shift
corresponding near-field signal at 6 m distance (probe distance corresponds to the feed longitudinal displacement of:
at the DTU-ESA Facility) was calculated for the two main 6mm/360º·0.25º = 0.004 mm = 4 µm. This result is
scanning schemes: θ-scan and ϕ-scan. The main effect of the significantly larger than the 1 µm obtained from GRASP
longitudinal feed displacement is the phase change of the simulations, since the effect of the near-field phase change
radiated signal proportional to the electrical length of the reduces after the near-to-far-field transformation.
displacement. Therefore, the near-field signals were then
modified such that a linearly-changing phase shift was added to VII. CONCLUSION
the near-field signal in each measured point so that the total A multi-band millimeter-wave VAST (mm-VAST) antenna
phase shift at the end of the measurement constituted the covering K/Ka/Q/V-bands is currently under development in
desired value, e.g. 1º. This error model corresponds to a linear collaboration between DTU and TICRA under contract from
displacement of the feed along its symmetry axis e.g. with the the European Space Agency. A detailed set of electrical and
changing temperature. The modified near-field signals were mechanical requirements was established to ensure maximum
then processed applying usual near-to-far-field transformation value of the mm-VAST antenna for facility comparison and
and the obtained far-field patterns with the induced errors were validation. The electrical baseline design is based on a single
compared to the reference far-field pattern. offset reflector antenna fed by a cluster of four Potter horns,
Since the developed mm-VAST antenna is expected to have one for each of the four frequency bands. A preliminary
the main beam pointing slightly off z-axis of the mechanical mechanical end thermal tolerance study has shown that the feed
coordinate system, the simulations were done for the main position should not change more than some 4-5 μm. It is
beam pointing along z-axis and also rotated by 5º from the z- planned to use a carbon fiber reinforced polymer, CRFP, for
axis of the measurement coordinate system along x-axis and manufacturing the support frame of the mm-VAST antenna in
along y-axis. In order to make proper comparisons (so that the order to achieve the required mechanical and thermal stability.
pattern cuts go through the main beam), the above rotations The detailed electrical and mechanical designs are currently on-
were then compensated by the reverse rotations before the going at DTU and TICRA, while manufacturing is planned for
comparisons. For each antenna orientation, the linear phase the winter 2014-2015.
shift was adjusted such that the achieved EES was at the level ACKNOWLEDGEMENT
of −60 dB. An example of the EES result together with the
pattern itself is shown in Fig. 5. It is seen from Fig. 5 that for This work is supported by the European Space Agency
the main beam pointing along z-axis the θ-scanning scheme is under the ESA Contract No. 4000109866/13/NL/MH.
very sensitive to the phase shift: with only 0.7º total phase shift
REFERENCES
the EES reaches −60 dB limit. On the other hand, the
sensitivity of the ϕ-scanning scheme (not shown here) is very [1] J. Lemanczyk and F. H. Larsen, “Comparison of near-field range
results,” IEEE Trans. Antennas Propag., vol. AP-36, pp. 845–851, Jun.
small: the total phase shift of 25º is necessary to reach the 1988.
−60 dB EES limit.
[2] C. F. Stubenrauch, et al, “International intercomparison of horn gain at
X-band,” IEEE Trans. Antennas Propag., vol. AP-44, pp. 1367–1374,
Oct. 1996.
[3] J. Fordham and M. Scott, “Antenna pattern comparison between an
outdoor cylindrical near-field test facility and an indoor spherical
nearfield test facility,” IEEE Antennas Propag. Mag., vol. 46, pp. 146–
151, June 2004.
[4] J. Lemanczyk, “Antenna standards and laboratories in Europe,” IEEE
Trans. Antennas Propag., vol. AP-44, pp. 347–350, Apr. 1995.
[5] S. Pivnenko et al., “Comparison of Antenna Measurement Facilities with
the DTU-ESA 12 GHz Validation Standard Antenna within the EU
Antenna Centre of Excellence”, IEEE Trans. Antennas Propag., vol. 57,
no. 7. pp. 1863-1878, July 2009.
[6] O. Breinbjerg, S.N. Pivnenko, O.S. Kim, J.M. Nielsen, "Recent
Advances in Antenna Measurement Techniques at the DTU ESA
Spherical Near-Field Antenna Test Facility", Proceedings of the URSI
Figure 5. VAST-12, z-axis pointing, θ-scan, 0.7º phase shift. General Assembly and Scientific Symposium, Beijing, China, 2014.
[7] C. Granet, T.S. Bird, “X-, Ku- and Ka-band Compact Feed Horns for
This behavior is explained by recalling that in the ϕ- Global-Earth Coverage from a Geostationary Orbit”, Proc. Asia Pacific
scanning scheme, the main beam (pointing along z-axis) is Microwave Conference, 3-6 Dec. 2000, Sydney, Asutralia.
scanned within the first few scans and thus rather large phase [8] T.G. Smith et al., ”An offset-fed 20/30 GHz dual-band circularly
changes are necessary to produce an error. In the θ-scanning polarized reflectarray antenna”, Proc. of the 7th European Conf. on
scheme, the main beam is crossed in every scan and thus the Antennas and Propag. (EuCAP), Gothenburg, Sweden, April 2013.
changing phase results in a discontinuity between the first and [9] K. Pontoppidan (Ed.), "GRASP Technical Description", TICRA, June
2008.
the last scans, which should be rather small for the given

View publication stats