The ESO NTT has been the first telescope whose enclosure is designed
to allow wind through-flow in the telescope volume (see section
above). It has also become a kind of a standard
reference for advanced telescopes in the 4-m class.
The same design of telescope and enclosure
with minor modifications has
been adopted for the 3.5-m GALILEO telescope presently in
construction for installation at the observatory of La Palma
(Canary Islands) and is considered
also for a future South-African telescope.
This enclosure is actually a real building, with several volumes and
rooms for telescope, instruments and auxiliary equipment which
rotates about the azimuth axis accompanying the telescope.
The telescope is housed in a kind of
corridor between two walls which separate the closed rooms for the
Nasmyth instruments which are air conditioned separately.
This corridor is closed on the front and upper sides
by 2 large upside-down-L shaped sliding doors while the back a wall includes
several louvers allowing through ventilation.
A semi-permeable windscreen can be raised across the
front side of the slit (fig. ).
Figure: The enclosure of the NTT with the stowable wind screen
along the front slit.
The tests were performed in the turbulent wind tunnel of the Technical University of Aachen with a 1:50 model of the NTT rotating building. The velocity measurements were taken at three locations, indicated in fig. 4.8. The tests were run with several venting configurations of the enclosure
The measurements covered the range of azimuth angles from 0 to
60
with the wind direction.
Figures 4.9 and 4.10 show the main results for
point TR45:
))
to the measured spectrum in the
inertial range. In this way the derived value of length scale will
best represent in parametric studies using the Von Karman expression
the actual energy density in the high frequency range most critical
for the telescope tracking performance. In the plot the length scale
is normalized with the slit width.
of the dynamic pressure term
, normalized with the term 
.
. Without windscreen (configurations O-O-O and
O-O-20%), this is associated with a
large increase of turbulence, particularly for azimuth angles between
10 and 40
where the turbulent fluctuations of
dynamic pressure created by the slit
are quite larger than in the free wind flow.
The turbulence length scale is respectively 0.37 and 0.48 of the
slit width at zero azimuth, which signifies the generation of high
frequency turbulence by the slit
edges even when the telescope faces the wind directly.
Fig. 4.11 shows the values of
the pressure spectral density at 6 Hz for a free
flow mean speed of 18 m/s, a good indicator of the wind
influence on the tracking performance as telescopes have generally
their first eigenfrequency between 5 and 8 Hz:
the amplification with respect to the free flow
conditions reaches one order of magnitude.
This effect is only corrected when the windscreen is raised to full
height across the front side of the slit, such that the mean flow
velocity, hence the pressure fluctuations on the telescope
(see equation ()) are sharply decreased.
On the basis of these
results a 20% permeability windscreen was integrated in
the final design of enclosure. Observation records of the NTT at La
Silla show that in most cases the operator is obliged to raise it
when the wind exceeds about 6 m/s, that is almost half of the
time of observation.
Figure 4.8: Drawing
of the model of the NTT building (dimensions are
full scale) showing the three measurement points.
[IMAGE ]
Figure 4.9: Flow
on telescope upper part for different venting
configurations - test results
Figure: Normalized rms of dynamic pressure on the telescope
upper part: 
Figure: Pressure spectral density at 6 Hz for a free flow wind
speed of 18 m/s.
For reference the free flow value (for L = 100 m) is 3.32 Pa
/Hz.
![[IMAGE ]](img248.gif){kind=link}