next up previous contents
Next: Dome seeing Up: Engineering criteria and Previous: Wind turbulence on

Wind buffeting on the primary mirror

The surface figure of the relatively thin primary mirrors of the new 8-m generation are quite sensitive to the turbulent pressure fluctuations caused by wind buffeting. For instance in the case of the VLT, the relationship (gif) leads to specify, for a wavefront error of 200 nanometers, a threshold for the admissible pressure fluctuation at 1.3 Pa. This is a very low value that will generally be reached with a wind speed close to the mirror of about 3 m/s.

From the practical engineering standpoint there are in principle three possible technical solutions to this problem:

  1. Passive solution through wind shielding.
    Here the enclosure must be capable of shielding the mirror from nearly all wind loading. This is the approach taken by the VLT project, where the relationships (gif) and (gif) will allow the designers to relate the airflow speed near the mirror with the pressure fluctuations, thereby providing the quantitative data for evaluating the optical performance of the mirror under wind loads and then for determining the best operating strategy for the venting devices (windscreen, louvers) of the enclosure. This topic will be further discussed in section 6.2 below.

    The other alternatives require the upgrading of "active optics" mirror support systems in order to provide a reaction capability of the mirror to fluctuating wind loads.

  2. Increasing the number of virtual fix-points.
    This approach has been taken in particular by the Gemini project (see page gif), where it is planned to provide the mirror with six fix-points instead of three. This will reduce the mirror deflections, hence the optical aberrations by a factor of about four.

  3. Increasing the bandwidth of the active optics control loop.
    We have seen that the peak frequency of the pressure fluctuations on a 8-m mirror is below 1 Hz in all the configurations tested. Therefore an extension of the active optics control loop up to at least 1 Hz, which appears technically feasible [Wilson 93], would achieve a sufficient dynamic correction of the effect of wind buffeting.
Clearly, while the first alternative is completely within the design domain of the enclosure engineer, the other two will require a concurrent approach, in which the development of improved active optics systems is driven by a parametric analysis of the amplitude and frequencies of pressure fluctuations on the mirror, performed along the lines illustrated in section gif.



next up previous contents
Next: Dome seeing Up: Engineering criteria and Previous: Wind turbulence on



Lorenzo Zago, zago@elgc.epfl.ch, Mon Nov 6 23:33:14 GMT+0100 1995