This phone meeting was attended by J.Bieging, J.Cheng, D.Emerson, M.Holdaway, J.Kingsley, J.Payne and D.Woody. The agenda for the meeting was:
(1) Gravity Performance: There now seems to be agreement that we understand the differences between the az/el (conventional) mount and the slantaxis mount with respect to the gravitational deformation of the reflector. This was clarified in David's e-mail of 15 July which gives the gravitational component of the surface rms as follows:
The conventional antenna is worst at the horizon and zenith with an rms of 11.4um, but decreases to 4.5um at ~40deg.
The slant-axis antenna is pretty flat achieving 9.9um at the zenith, horizon and ~45deg and minima of 9.1 at ~23 and 68deg.
Thus the difference between the two is small. These deformation models can be used in the Ruze equation to estimate the difference in sensitivity for the two mounts as a result of surface gravitational deformation. At 850 Ghz, taking an unweighted average of the aperture efficiency over the elevation angle range 0 to 90 deg, the slantaxis design is less sensitive than the az/el mount by a factor 0.96. If a reasonable weighting function is used to account for the fraction of observing time expected for each elevation angle, this factor becomes 0.93.
David makes this additional point in his 15 July e-mail: It should also be noted that the conventional antenna is much less risky. If our deformation models are wrong by a factor of two then you can still tune the conventional antenna for better than 12um rms over most of the middle of the zenith angle range. Whereas the slant-axis with min-max tuning would be at ~19um rms over the full range. Tuning the slant-axis to be perfect at 45deg would give a very narrow elevation range of acceptable performance.
(2) Close Packing Performance: We need to understand the significance of the apparent difference in the close packing performance of the two mount designs. As the two designs sit at present their close packing properties are as follows:
Conventional Slantaxis
No active protection,0 deg elev limit 1.43 D 1.6 D
No active protection,30 deg elev limit 1.35 D 1.8 D
Active protection, all antennas pointing
in same direction 1.08 D 1.11 D
Either design is good enough if we provide active anti-collision protection, so it seems that there is only a significant difference between the two designs if we can improve the conventional design to make it good enough without active protection (there seems little hope of making the slantaxis acceptable without active protection). In MMA Memo 155 Mark shows that close packing of 1.48 D at 0 deg and 1.30 D at 30 deg is probably adequate for the astronomy. It would be desirable, but not necessary, to achieve 1.30 D at 0 deg in which case we would not need the complication of a variable elevation limit. In his e-mail of 12 July David shows that we may be able to achieve 1.30 D at 0 deg. Some of the changes that we would have to make to the design to achieve this include reducing the depth of the reflector backup structure, changing the F/D ratio, moving the quadrupod support points in from the edge of the dish slightly and minimizing the depth of the subreflector moving mechanism. We need to answer the question "can we make the conventional design good enough without the need for active protection?".
(3) Status of the Mount Comparison Report: At the face-to-face meeting of the working group in May 96 there was a good consensus that the conventional design was the best design for the MMA. This conclusion was based on a comparison of the advantages and disadvantages of the two mounts. The conventional antenna had problems with pointing and the slantaxis had problems with gravitational deformation, close packing and the risk of a design previously unused for radioastronomy. Since that meeting the strength of the conclusion has weakened somewhat because the difference in gravitational performance is much less than originally thought and it is not yet clear whether the difference in close packing is significant. However, it still seems appropriate to concentrate on the conventional design at the present time because we must answer the question "how good can we make the pointing of the conventional antenna". The current conventional design has an average pointing error in 9 m/s wind of 3.5 arcsec vs 1.2 arcsec for the slantaxis. Note that the average in this case is over all antenna azimuth and elevation angles and, whilst the comparison between mounts is probably reliable, the absolute value of the pointing error is uncertain due to the difficulty of accurately calculating wind load forces on the complex reflector structure. The report will detail what we currently understand about the differences between the two designs and will recommend that we concentrate on the conventional design at least until we can answer the question of pointing performance. Jinquan has made a first draft of the report and Peter is working on a second draft which will be distributed to the working group for comment.
(4) Status of Conventional Design: Jingquan and Jeff are currently working on the receiver cabin and the support ring between the cabin and the backup structure. They will try and move the elevation axis forward which will improve both pointing and close packing performance. To improve pointing performance we must investigate an improved (probably asymmetric) yoke design and an independent support for the elevation encoder. To improve close packing performance the modifications listed at the end of section (2) above must be studied.
Last modified 12 August 96
mholdawa@nrao.edu