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

net/publication/238581414

Thermal and Wind Effects on the Azimuth Axis Tilt of the ASTE 10-m Antenna

Article · January 2007

CITATIONS READS

16 386

4 authors, including:

N. Ukita Bungo Ikenoue

National Astronomical Observatory of Japan National Astronomical Observatory of Japan
164 PUBLICATIONS   1,927 CITATIONS    41 PUBLICATIONS   561 CITATIONS   

SEE PROFILE SEE PROFILE

Masao Saito
National Astronomical Observatory of Japan
259 PUBLICATIONS   3,463 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Atacama Large Millimeter and Submillimeter Array View project

ALMA Construction View project

All content following this page was uploaded by N. Ukita on 05 February 2014.

The user has requested enhancement of the downloaded file.

Publ. Natl. Astron. Obs. Japan Vol. 10. 25 – 33 (2007)

Thermal and Wind Effects on the Azimuth Axis Tilt

of the ASTE 10-m Antenna

Nobuharu UKITA, Hajime EZAWA, Bungo IKENOUE, and Masao SAITO

(Receved April 27, 2007; Accepted June 21, 2007)

Abstract
The azimuth axis tilt of the ASTE 10-m antenna induced by thermal and wind loadings was investigated with a
dual-axis inclinometer on the azimuth axis, along with thermometers on the pedestal and yoke structures and an
ultrasonic anemometer on a nearby weather station. The dependences of the inclinometer zero-point offsets
against temperature of the device, temperature gradients in the pedestal and yoke structure were obtained for the
measurements over 11 months during the antenna being parked at its home position (azimuth angles = ±180
degrees, an elevation angle = 60 degrees) under wind velocities < 8 m s–1. The temperature dependences of the
zero-point offsets were found to be 1.24 and -0.46 arcseconds/degree, and were close to those obtained with an
independent method. The azimuth axis tilts due to the temperature difference between the two opposite sides of
pedestal walls were found to be about 1.1 and 1.7 arcseconds/degree, and consistent with 1.5 arcseconds/degree
estimated with a simple model. The residual axis tilt of the whole samples after removal of the temperature depen-
dences shows dependence against overturning moment estimated from the wind data. The stiffness of the antenna
structures between the yoke base section and the ground was estimated to be 5.3 and 3.4 GNm/rad using the
observed tilts in two directions; and were smaller than 6.0 GNm/rad from a mechanical model prediction. Based
on these field experiments, we discuss the improvements and limitations of pointing performance with the incli-
nometer metrology system.

Keywords: radio telescope: pointing, inclinometer, thermal load, wind load

1. Introduction ical misalignments and gravity deformation that can be cor-

rected with a simple pointing model with 8 to 20 terms (e.g.,
Pointing accuracy of a radio telescope is one of key per- Greve et al., 1996; Ukita, 1999; Ukita et al., 2004; Mangum
formances that determine the quality and throughput of data, et al., 2006). Errors due to friction, servo errors in drive sys-
which in the end influences research productivity of the tele- tems, and rapidly changing wind turbulence are non-repeat-
scope. It is desirable that blind pointing errors on the whole able. Pointing errors due to thermal and static wind loadings
sky do not exceed about one fourth of a telescope beam- are in-between. Thermal deformations are, in principle, pre-
width for pointing checks/corrections prior to observations, dictable from temperature measurements of the antenna
and that tracking errors are less than one tenth of the beam- structure with the aid of finite element model (FEM) calcula-
width to keep a sensitivity loss within 3 percents for a com- tions (e.g., Bremer and Peñalver, 2002). The static wind
pact object for an hour or two. The 10-m antenna of the effects can also be estimated using coefficients derived from
Atacama Sub-millimeter Telescope Experiment (ASTE) pro- wind tunnel and/or field experiments. These two effects are
ject of Nobeyama Radio Observatory* is an experimental slow in time and can also be corrected with a telescope con-
telescope for sub-millimeter observations (Ukita et al., 2000; trol computer.
Ezawa et al., 2004) at a windy site. Because the antenna is The structure surface of the ASTE antenna is protected
exposed to wind and sunshine, it is a tough task to achieve with thermal insulation boards to minimize the thermal
pointing performance requested for the antenna whose beam- deformation. There will be, however, inevitable thermal
width is 7 arcseconds at 950 GHz. effects on the antenna structure. For example, a tilt of a steel
There are two types of pointing errors, repeatable and square box amounts to 2.3 arcseconds when it has tempera-
non-repeatable errors. The repeatable errors include mechan- ture gradient of one degree across the two opposite sides of
its walls. Drag force of head wind of 20 m s–1 on a 10-m dish
*Nobeyama Radio Observatory is a branch of the National is about 17 kN at the site. It is estimated that this loading on
Astronomical Observatory of Japan, an inter-university research insti-
the ASTE antenna (113 kNm) makes the yoke base structure,
tute operated by the Ministry of Education, Culture, Sports, Science,
and Technology. the azimuth bearing structure, the antenna pedestal, fastener
bolts, and the concrete foundation bent by a total of about 4
26 Nobuharu UKITA and others

arcseconds. These estimates clearly suggest that we need

metrology systems to cope with the inevitable deformations
of the antenna mount structure. Many telescopes have used
high precision inclinometer(s) as a metrology device for
improvements of pointing performance (e.g., Kibrick et al.,
1998; Ukita 1999; Peñalver et al., 2000; Parker and Srikanth,
2001, Prestage et al., 2004). The ASTE antenna also has a
dual-axis inclinometer on the azimuth axis to measure the
axis tilt and correct the pointing offsets accordingly in real-
time (Ukita et al. 2000).
This report describes one of such metrology systems to
achieve the ultimate goal of the pointing performance of the
ASTE 10-m telescope. We investigate whether the incli-
nometer consistently detects the azimuth axis tilts induced
by thermal loads measured in the field, and whether it has
long term stability good enough for our pointing goal. We
also look into wind load effects on the mount structure.
Section 2 describes our experiment system and measurement Figure 2. Diurnal variations of ambient temperature (upper) and
data for a year. Section 3 shows our analyses of both thermal wind velocity (lower) measured from February to December, 2005.
The white dashed lines indicate local noon. The antenna location is
and wind loading deformations. Section 4 discusses pointing
23 deg south, 4h30m east.
performance improvements by the inclinometer metrology
system and its limitation. desert, northern Chile. Figure 2 shows diurnal variations of
ambient temperature and wind velocity. During night the
2. Equipment ambient temperature is stable and wind velocity is low (a
typical mean velocity of 4.5 m s–1), and during daytime tem-
The ASTE 10-m antenna has an alt-azimuth mount. perature changes are large and wind velocity is high (typical
Figure 1 displays its steel mount structure; its pedestal has an mean velocities >10 m s–1 in the afternoon). A prevailing
average diameter of 2.7 m and a height of 1.8 m and the ele- wind direction is west, and a mean air pressure is 570 hPa.
vation axis is supported by its yoke arms at a height of 6.5 m We used a dual-axis inclinometer (Nivel20, Leica) with
from the ground. The antenna is located at Pampa la Bola a digital output port (RS232C). It incorporates a fluid mirror
(23 deg south, 4h30m east, and altitude 4,800 m) in Atacama and a position sensitive diode to measure tilts in two axes. It

Figure.1 The pedestal and yoke structure of the ASTE 10-m antenna. Star marks are the location of the inclinometer, dots indicate the temper-
ature sensors that are installed on inner surfaces of the structure. The antenna is at Az = ±180 degrees (South). The coordinate system used in
this report is also shown.
本文（欧文）10-1,4-03 07.12.11 7:19 PM ページ27

Thermal and Wind Effects on the Azimuth Axis Tilt of the ASTE 10-m Antenna 27

provides a resolution of 0.001 mrad (0.2 arcseconds) and a 3. Analysis

dynamic range of ±1.5 mrad (±5 arcminutes). It has an
output time constant of about 1 second. The data are logged The data we handled in this report are those logged
every second. It is located on the azimuth axis in the yoke when the antenna was parked at its home position angle of
structure (see figure 1) to avoid the centrifugal force of the Az = ±180 and El = 60 degrees for intervals longer than six
antenna azimuth movement. Its zero-point offsets vary with hours. There were 101 such occasions in the period from
the device temperature. According a specification sheet, they February to December, 2005. To reduce data samples, we
shall be less than 0.005 mrad/degree (1 arcsecond/degree). averaged the data from the inclinometer and anemometer for
They can be measured from the average of readouts during a one minute, which resulted in 92,000 samples. The ther-
full turn of 360 degrees in azimuth. Their temperature mometer data were interpolated for each sample point. In
dependences can be determined from a series of such mea- figure 3, the measured individual tilts are plotted as dots to
surements at different device temperatures (e.g., Ikenoue et show diurnal variations. Typical amplitudes of the diurnal
al., 2005). They are 1.29±0.15 arcseconds/degree in and variations in and are 10 and 6 arcseconds, respec-
-0.29±0.09 arcseconds/degree in (see Appendix). tively. It seems that they can be interpreted in terms of the
zero-point offset variations and temperature gradients in the
antenna structure, based on the pre-considerations and data-
inspections mentioned in sections 1 and 2.
In addition, it should be noticed that in figure 3 there
are both thin and burred traces in time, which seems to be an
indication of a wind effect. Therefore we temporally set a
wind velocity threshold of 8 m s–1 in the following thermal
effect study, which results in about 54,000 samples. This
threshold value is a compromise to have a certain high num-
ber density of sample points in daytime where temperature
variations are large and to minimize the wind load effects.

3.1 Thermal Effects

By the above selection of the antenna position and wind

velocity, it is now possible to make a simple model to
describe the azimuth axis tilts. We use following empirical
Figure 3. Diurnal variations of tilts (upper: in ; lower: ) equations:
measured from February to December, 2005. The antenna was at its
, (1)
home position of Az = ±180 deg. and El = 60 deg.
and
The ASTE antenna has about 170 thermocouple ther- , (2)
mometers (T-type) on its antenna structure, and 14 of them where T0 is the inclinometer temperature,
at the pedestal and yoke structure (see figure 1). The outputs and (=T 4 – T 2) are temperature gradients of the
from the thermometers have been recorded at 3-minutes pedestal structure, (=T5 – T6) and (=T14 – T13)
intervals through a data logger (THERMODAC-6, Etodenki are temperature gradients in the yoke structure.
Co., Ltd) with a resolution of 0.1 degrees. We have recognized an additional problem. The incli-
A 2D ultrasonic anemometer (WindObserver II, Gill nometer outputs jump in a discontinuous manner when
Instruments Ltd) is installed atop of a weather tower at a changing from an active drive state by the electronic con-
height of 6 m, comparable to that of the elevation axis of the trolled drive motors to a parking state engaged by mechani-
antenna. The tower is located in the SW direction (Az = -134 cal brakes. The jumps are typically a few arcseconds and
deg.) and at 35 m from the antenna. The data are logged at 4 non-repeatable. Therefore we take the difference approach of
Hz. data analysis for these equations. We search for epochs at
which the wind velocity is a minimum in individual data
sets. Figures 4 and 5 display these difference values thus
本文（欧文）10-1,4-03 07.12.11 7:19 PM ページ28

28 Nobuharu UKITA and others

Figure 4. (a) Elevation tilts ( ) corrected for tilt jumps due to brake operation of the drive motors (see the text); (b) the inclinometer tem-
perature (T 0); (c) temperature difference between front- and rear-side of the pedestal (T1 – T3); (d) temperature difference between front- and
rear-side of the yoke base (T5 – T6) for the antenna position of Az = ±180 and El = 60 degrees and wind velocities < 8 m s–1. Note that the ver-
tical scales are different among the panels.

obtained for the tilt, the inclinometer temperature, and the namely about 1 degree in the morning and about –1 degree
temperature gradients in the pedestal and yoke structures. in the evening (figure 5c). On the contrary, the south-north
The inclinometer temperature variations are smooth temperature gradient variations (figure 4c) show large ampli-
(figures 4b and 5b), and have large time lags with respect to tudes of about 3 degrees and mimic the ambient temperature
the ambient temperature (figure 2a). The amplitudes are typi- variations. This is because the southern part of the pedestal
cally 5 degrees. The east-west temperature gradient varia- has the azimuth gear boxes and its surface is not protected by
tions in the pedestal clearly show the effect of sunlight, insulator boards. The temperature of this part follows closely

Figure 5. Same as figure 4, but for (a) cross-elevation tilts ( ); (c) temperature difference between east- and west-side of the pedestal (T4 –
T2); (d) temperature difference between right- and left-side of the yoke (T14 – T13).
 Thermal and Wind Effects on the Azimuth Axis Tilt of the ASTE 10-m Antenna 29

Table 1. Temperature dependence of tilt.

the ambient temperature. Figure 4d shows the front-rear tem- for instrument boxes which make shadows on the left half of
perature gradient variations of the yoke, which is also the the rear faces of the yoke, which in turn produces a tempera-
effect of sunlight, namely about –2 degrees near noon. ture gradient along the elevation axis direction during day-
We have made least squares fits for the differential tilts, time. Because in the first step of our analysis the samples
and have obtained coefficients for these variables (table 1). mostly consist of those during night (low wind velocities),
The uncertainties of coefficients have been estimated from this thermal effect due to the sunlight does not appear clear-
the cases of wind velocity thresholds of 7 and 9 m s–1. The ly.
temperature dependences of inclinometer zero-point offsets The coefficient for the front-rear temperature gradient
in , C1=1.24 ±0.02 arcseconds/degree, and in , C 5= in the yoke structure, C3, was 1.54 arcseconds/degree (table
–0.46±0.05 arcseconds/degree are close to 1.29±0.15 and 1), and is larger than that estimated from a simple model (1.0
–0.25±0.09 arcseconds/degree, respectively, obtained from arcseconds/degree). These suggest that the yoke structure
the independent method. The azimuth axis tilt due to the has a scale length of temperature gradient shorter than the
temperature difference between the two opposite sides of distances between the sensors. We need more thermometers
pedestal walls is estimated to be about 1.5 arcseconds/degree to reveal temperature distributions in the yoke structure.
with a simple model. The observed coefficients were 1.07 Therefore it might to be not useful to make comparisons of
and 1.73 arcseconds/degree. the coefficients from the measurements and simple models
that assume a linear gradient across the structures.

3.2 Wind Effect

Wind loads on a parabola dish can be estimated using

the table of force coefficients obtained from wind tunnel
experiments (Levy, 1996). The f/D ratios of the dish used in
the tunnel experiments and of the ASTE antenna are 0.33
and 0.35, respectively; and close each other. We have to
notice, however, the shape difference between a thin parabo-
Figure 6. Residual tilts in after the adjustments of azimuth tilt la of the model (like a satellite antenna for home use) and the
jumps by brake operations against temperature gradient in the yoke
ASTE main dish. The back-up structure of the ASTE main
structure, (see the text). The thin line shows a lin-
dish has a thickness of 1.8 m at its center, comparable to the
ear fit.
depth of front surface profile of the reflector; in other words
The tilts of the yoke structure show complex behav- the projected area viewed from its side is twice that of a thin
iours. After the corrections with the coefficients in table 1 parabola. We used the side force coefficients modified
and the adjustments of offsets due to the jumps of brake accordingly. For a receiver cabin we simply assumed that Cd
operations, which will be described in section 3.2, we have is 2 (Simiu and Scanlan, 1996). Table 2 gives overturning
searched for additional correlations between the residual tilts moments by wind of 10 m s–1 on the ASTE antenna for 13
in and and differences among temperatures at the 14 directions. We can estimate the stiffness of the antenna
locations for the entire data samples. There have been found mount by comparing observed tilts with wind loads calculat-
no significant correlations in the . We have, however, ed with table 2. We use following empirical equations:
found a correlation between the residual tilts in and the , (3)
difference between the rear face of the yoke base, T6, and the and
rear face of right yoke arm, T10 (figure 6). A possible reason , (4)
is the following. The asymmetric layout of a rear platform where Mx and My are overturning moments.
30 Nobuharu UKITA and others

Table 2. Overturning moment by wind of 10 m s–1.

We have encountered again the problem of the incli- 3.3 Contents of Tilt Variations
nometer output jumps due to brake operation. In order to
estimate the jumps of the individual data sets, we have made In summary we have untangled the observed inclinome-
least squares fits to each data set with equations 3 and 4 to ter outputs with six terms of sources. The large variations
find constants of C10 and C12. These constants were subtract- and scatters seen in figures 3a and b are reduced those shown
ed from the individual data sets. Figures 7a and c display the in figures 7a and c after the removal of the thermal loading
tilts of all the 92,000 samples after the corrections of thermal and brake operation effects, and to the final unknown residu-
effects and these offset jump adjustments. als shown in figure 8a and b after the elimination of the wind
Figures 7b and d show the tilts against the overturning loading effect. We have calculated their standard deviations
moments by wind. We have made least squares fits to the and listed in the second and last columns of table 3. We have
whole samples. The inverse of C9 and C11 give the stiffness evaluated them in the time ranges from UT=3h to 9h (night)
of the antenna structures between the yoke base section and and from UT=15h to 21h (daytime). We have also calculated
the ground. They are 5.3±0.3 and 3.4±0.1 GNm/rad in the standard deviations for the individual sources of variations
two directions parallel and perpendicular to the elevation (from the third to 8th columns of table 3). The biggest con-
axis, respectively. These values are smaller than 6.0 tributor was the temperature variation of the inclinometer
GNm/rad from a model calculation. Angles of bend by a itself. The second was the jumps due to brake operation,
loading of 113 kNm on the yoke base, the azimuth bearing, which is not relevant to the observing performance. The
the antenna pedestal, fastener bolts, and the concrete founda- third was the tilt in due to the temperature gradient
tion are 0.2, 2.1, 0.7, 0.8, and <0.1, respectively; and a total between the front- and rear faces of the yoke base.
of about 4 arcseconds.

Figure 7. Residual tilts in and (a, c) and their dependences against wind loads Mx and My, respectively (b, d). The thin lines show lin-
ear fits.
 Thermal and Wind Effects on the Azimuth Axis Tilt of the ASTE 10-m Antenna 31

Table 3. Contents of tilt variations.

Ironically, most of the variations were induced by the

inclinometer temperature variation. One may propose to
keep the temperature of inclinometer constant with an addi-
tional device(s) (e.g., a controller with heater). Such active
approach might produce a thermal deformation of the struc-
ture near the inclinometer and it is no longer the reference
plane for the inclinometer. Our passive approach is to leave
it in a simple temperature distribution and to minimize local
deformations. The ASTE antenna has no heat sources such
as electric devices inside the pedestal and yoke structures,
except the azimuth angle encoder and the inclinometer. This
is one of our antenna design concepts as an experimental
telescope. The motivation of this report is based on this
approach. Therefore it is essential to calibrate the device
accurately.
We estimate a required calibration accuracy from our
field data. The standard deviation of the inclinometer tem-
peratures during the 11 months was 3.5 degrees. If we tem-
porally assume that the error allocated in a blind pointing
Figure 8. Residual tilts in and after corrections of wind error budget for the inclinometer itself to be 0.5 arcseconds,
load effects.
the calibration accuracy should be about 10 %, namely 0.5
arcseconds divided by the coefficient of 1.24
4. Discussion arcseconds/degree and by 3.5 degrees. The typical slope of
inclinometer temperature variation in figure 4b was 0.67
Our inclinometer is a key metrology device for the degrees/hour. If we assume that the error allocated in a track-
antenna pointing performance. Our analysis in the previous ing error budget for the inclinometer itself to be 0.2 arcsec-
section seems to suggest that the inclinometer fairly consis- onds, which is equal to the inclinometer output resolution,
tently detects the azimuth axis tilts and that it has long term the calibration accuracy should be again about 10 %, namely
stability far better than 1 arcsecond (figure 8). These results 0.2 arcseconds divided by the coefficient of 1.24 arcsec-
are encouraging and clearly deserve further studies in detail onds/degree, by the temperature slope of 0.67 degrees/hour,
in the following pages. and by 2 hours of tracking time span. Since the discrepancy

Table 4. Improvement and limitation of pointing performance with the inclinometer system.
32 Nobuharu UKITA and others

between our two measurements with the independent meth- pointing goal.
ods was within 10 %, we conclude that the required calibra- Finally we discuss about the wind load. The measured
tion accuracy has been achieved. stiffness was smaller than the model prediction. It is difficult
The following three paragraphs discuss the improve- to estimate the uncertainties of force coefficients (Levy,
ment and limitation of pointing performance with the current 1996), and it seems that we should be satisfied with our
inclinometer metrology system. The inclinometer detects results within a discrepancy factor of 2. It is highly desirable
both a large scale and local deformations below the yoke that we make dedicated experiments with an additional
base structure, namely in the yoke base itself, azimuth bear- device(s) under a specific condition to make it clear what is
ing, pedestal and concrete foundation. One of the key roles the source of this discrepancy. A proper antenna-wind con-
of the inclinometer is to eliminate effects due to mechanical figuration is an attack angle of zero degree, e.g., when the
deformations of the azimuth bearing, such as its non-repeat- wind direction is –90 degree, the antenna is at Az = –90 deg.
able runout, that cannot be described with a simple pointing and El = 0 deg. Because the configuration is simple and
model. Aside this, we look into the advantages and disad- there have been a number of tunnel experiments, the coeffi-
vantages of the inclinometer system. The advantage is, of cients for this configuration are most reliable. Possible loca-
course, to detect a tilt in a large scale, namely the azimuth tions of the additional inclinometer(s) are on the concrete
axis tilt. The disadvantage is that if there are local deforma- foundation and beneath the azimuth bearing.
tions in the yoke base irrelevant to the true axis tilt, the In summary, we have investigated the azimuth axis tilt
metrology system makes a false correction of pointing. of the ASTE 10-m antenna induced by thermal and wind
Here we assume that in our experiments the true axis loadings with a dual-axis inclinometer. We have interpreted
tilts are those detected with linear fits in table 3 (fourth to the observed tilts with six terms of sources (table 3). Most of
seventh columns), and the false tilts are the residuals in fig- the tilt variations were of thermal origin. We have revealed
ure 8a and b (the last column of table 3). We have calculated the tilt due to static wind loadings and derived the stiffness
standard deviations of improvements and false corrections of the antenna mount structure (figure 7). We have discussed
for the all-sky pointing and 2-hours tracking (table 4). We the improvements and limitations of pointing performance
have evaluated them in the time ranges from UT=3h to 9h with the current inclinometer metrology system.
(night) and from UT=15h to 21h (daytime) and for the case
of an elevation angle of 60 degrees. Table 4 shows that the The authors would like to acknowledge the ASTE team
system gives improvements in the all-sky pointing opera- members for their great efforts to the ASTE project. The
tions, but no significant net-improvements in the 2-hours authors also thank the NRO staff for their encouragement
tracking. If we assume that the error allocated in a blind throughout the work.
pointing error budget and the error allocated in a tracking
error budget for the false correction to be 0.5 and 0.2 arcsec- Variation of Inclinometer Zero-Point Offsets
onds, respectively, the current system meets this requirement
during night but not during daytime. These investigations Our dual-axis inclinometer (Nivel20, Leica) has tem-
suggest that we need a countermeasure against the local tem- perature dependences of zero-point offsets against its device
perature irregularity in the yoke base to achieve the pointing temperature. One method to obtain zero-point offsets is our
performance imposed on the ASTE antenna. method described in our paper, and another is to measure the
Although we have discussed only the deformations average of readouts during a full turn of 360 degrees in
between the yoke base and the ground, we should notice that azimuth. Their temperature dependences can be determined
it is essential to estimate the angular position of the elevation from a series of measurements at different device tempera-
axis. The relative angular position between the elevation axis tures (e.g., Ikenoue et al., 2005). Figure 9 shows such 190
and the inclinometer should be found with an accuracy nec- measurements made repeatedly on several days in the period
essary as the metrology system. Not only local deformation 2004 April 25 to June 9. It takes only 9 minutes for a single
near the inclinometer but also deformations of the two yoke measurement. We have made least squares fits, and have
arms should be estimated with high accuracy with a high obtained coefficients in the following empirical equations:
number density of sensors. Therefore we have recently
installed about 80 additional thermometers in the yoke arm , (5)
and base structures as a new step to achieve its ultimate and
 Thermal and Wind Effects on the Azimuth Axis Tilt of the ASTE 10-m Antenna 33

, (6) Ezawa, H., Kawabe, R., Kohno, K., and Yamamoto, S. 2004,
where T0 is the device temperature of the inclinometer. The “The Atacama Submillimeter Telescope Experiment
uncertainties in parameters are given in 3 . (ASTE)”, Proc. SPIE, 5489, 763-772.
Greve, A., Panis, J.-F., and Thum, C. 1996, “The Pointing of
the IRAM 30-m Telescope”, Astron. Astrophys. Suppl.,
115, 379-385.
Ikenoue, B., Ukita, N., Saito, M., Ezawa, H. 2005,
“Performance evaluation of an ALMA 12-m prototype
antenna with an optical pointing telescope system ”,
Report Natl. Astron. Obs. Japan, 8, 111-123 (in
Japanese).
Kibrick, R.I., Robinson, L.B., Wallace, V., Cowley, D.J.
1998, “Tests of a precision tiltmeter system for measur-
ing telescope position”, Proc. SPIE, 3351, 342-353.
Levy, R. 1996, “ Structural Engineering of Microwave
Antennas”, IEEE Press, Piscataway, NJ.
Mangum, J.G., et al. 2006, “Evaluation of ALMA Prototype
Antennas”, Publ. Astron. Soc. Pacific, 118, 1257-1301.
Parker, D.H., Srikanth, S. 2001, “ Measurement system for
the Green Bank Telescope”, Antennas and Propagation
Society International Symposium, 2001. IEEE, 4, 592 -
595.
Peñalver, J., Lisenfeld, U., Mauersberger, R. 2000, “Pointing
with the IRAM 30m Telescope”, Proc. SPIE, 4015, 632-
640.
Prestage, R. M., Constantikes, K. T., Balser, D. S., Condon,
J. J. 2004, “The GBT precision telescope control sys-
tem”, Proc. SPIE, 5489, 1029-1040.
Simiu, E. and Scanlan, R.H. 1996, “Wind Effects on
Structures”, John Wiley & Sons, Inc., Hoboken, NJ.
Figure 9. Variations of inclinometer zero-point offsets against its Ukita, N. 1999, “Thermal Effects on the Pointing of the
device temperature Nobeyama 45-m Telescope”, Publ. Natl. Astron. Obs.
Japan, 5, 139-147.
References Ukita, N. et. al. 2000, “NRO 10-m submillimeter telescope”,
Proc. SPIE, 4015, 177-184.
Bremer, M. and Peñalver, J. 2002, “FE model based inter- Ukita, N. et. al. 2004, “Design and performance of the
pretation of telescope temperature variations”, Proc. ALMA-J prototype antenna”, Proc. SPIE, 5489, 1085-
SPIE, 4757, 186-195. 1093.

View publication stats