Two important results were mentioned above:
).
Recalling that existing active optics systems, such the one installed on the NTT have a bandpass of about 1/30 Hz, [Wilson 89] analyzes in some detail the possible implementation and limitations of improved active optics systems with an extended bandpass up to 1 Hz and eventually 10 Hz. Various technical solutions are proposed and it is concluded that such systems are technically feasible. It is therefore interesting to elaborate the test results obtained during this work also in terms of the peak frequencies of the pressure fluctuations on the mirror.
Fig. presents a
scatter plot of peak frequencies versus the velocity ratio
at an 8-meter primary mirror inside the cylindrical
enclosure.
The data include all test configurations with varying
azimuth angles and venting conditions and are here only separated
according to the pointing zenithal angle.
Recalling the relationship () and that the
test wind speed was
about 12 m/s, we derive a range of turbulence length scales for the
pressure fluctuations between 0.9 and 3.6 m (full scale).
With respect to the mirror diameter the length scale range is then
During the mirror dummy tests at the NTT, the wind speed was not measured close to the mirror, so that a similar analysis is unfortunately not possible. However several power spectra of pressure fluctuation measurements were evaluated ([Noethe 92]) in which the peak frequency was within the range 0.2 to 2 Hz, which is exactly the same (in view of the 1:2 geometrical scale) as the range recorded in the cylindrical enclosure. This confirms also with respect to the frequency characteristics the general conclusion already drawn for modes and amplitudes of wavefront fluctuations: wind buffeting effects on the mirror depend first on the geometrical scale and only to a smaller extent of particular venting conditions.
We can then pragmatically conclude an upper bound for the worst case peak frequency of pressure fluctuations on a primary mirror, which is quite independent on the particular pointing angle or venting configuration:
This result, which stipulates that the peak frequency of significant wind buffeting effects on a large telescope mirror will in no case exceed the range of 1 to 2 Hz, will hopefully encourage the development of active optics systems into a frequency bandwidth which should be realistically achievable with present state technology.
Figure: Scatter plot of peak frequencies versus the velocity ratio
. The frequency data were obtained by visual interpretation
of the power spectra reported in the test report [DMI].