There is little effect from wind on the performance of 4-m
class telescopes
enclosed in large, oversize domes like the CFHT or the ESO 3.6-m
(see chapter 
). In fact one of the main design
criteria of those domes was the prevention of any wind loading on the
telescope by leaving a large distance between the telescope tip and
the slit and including movable windscreens to close those sectors of
the slit that are not in the field of observation. The only wind
tunnel study performed for these domes concerned static load
measurements on a model of the CFHT telescope ([ENSAM])
which concluded that static wind loads were indeed negligible
on the telescope.
More recently, wind tunnel tests were performed on a model of the
hemispherical dome
of the 10-m Keck telescope [Kiceniuk],
aimed at evaluating the airflow patterns and speeds inside the dome.
After that the MMT showed the favorable effects of wind on seeing, several aerodynamics studies were performed on enclosure design concepts in which the telescope volume was flushed by the wind flow. Worth mentioning are some water tunnel tests performed by Japanese and US teams searching for the best arrangement of venting openings in enclosures of hemispherical, cylindrical and various polyhedral form ([Ando], [Siriluk]).
We note here that the value of all these tests resides mainly in a qualitative evaluation of the flow patterns surrounding the telescope, since the similarity conditions were not respected sufficiently for drawing accurate quantitative results, generally because of the limitations of the test facilities utilized.
Also, these studies generally overlooked the importance of the
effect of the slit induced turbulence, first put in evidence by the
author
[Zago 85] in a preliminary analysis of the wind tunnel
measurements of the NTT building. At a time when venting a telescope
was reported as the newest and best solution to fight the dome seeing
problem
(see page ), we pointed out that in the
context of the new projects for 8-m telescopes
it could also affect negatively the telescope performance in two
respects:
acting on the telescope.
It may be useful to underline two aspects that make the subject
peculiar with respect to more conventional wind engineering.
The first aspect is the extreme smallness of critical
structural deflections that must be evaluated and ultimately minimized
by proper engineering.
With an aimed guiding accuracy of 0.3 arcsec rms
, the tip of an
8-m telescope shall keep its deflection under wind
to less than 20 
m rms.
Therefore optical telescopes are not operated with wind speeds exceeding 80 or
100 km/hr, as in these conditions the enclosure is closed.
One of the objectives of this research
is the determination of the relatively low wind loads acting
on a telescope in a way that is both accurate and
suitable for input in parametric studies of the telescope performance.
Analogously, the quality of the large primary mirrors starts
deteriorating when deflections are greater than 200 nanometers, and the
aberration depends strongly on the modal shape of the loading.
The second peculiar aspect is consequent to the fully automated active control of guiding and mirror support. These systems have allowed a tremendous progress in maintaining the accuracy of the telescope over the exposure times of the observations but are able to compensate external loads only within a certain frequency bandwidth. Therefore one is interested in particular in the accurate characterization of the high frequency component of wind load, even when this constitutes a small part of the overall load.