By the time you read this, we will have had a busy summer shutdown at the Mayall 4-m telescope, which will have gotten us a jump-start on several major initiatives in our continuing project to improve the delivered image quality (DIQ) at the 4-m. This article is intended to summarize where we are and where we expect to be shortly.
Major improvements began a decade ago by installing dome fans and cooling the horseshoe bearing oil. More recent changes have included the cooling system for the primary mirror and installing dome ventilation, in addition to the many minor improvements.
Over the past year we have completed an analysis of the behavior of DIQ with mirror temperature, and what thermal changes we have on a "typical" night. (Undergraduate Joshua Pearson undertook much of this horrendous task.) This work showed that during 90% of the nights the temperature drop on Kitt Peak is < 1.5ºC; on 80% of the nights it is a fraction of a degree, measured from an hour after sunset to dawn. We also found that the previous night's low is an excellent predictor of the next night's low.
There remains a good correlation of the DIQ with how far the primary mirror
is out of equilibrium, in accord to our expectations from the Racine et al.
study of the CFHT thermal environment (1991 PASP, 103, 1020). To summarize
this simply: the range of Tm (= Tmirror - Tdome) corresponding to ranges of
DIQs are:
< 0.80" -2.1ºC to +1.5ºC
0.80-1.10" -4.4ºC to +2.4ºC
1.10-1.50" -4.4ºC to +3.1ºC
This analysis suggests a reasonable goal would be 2.5ºC to +1.5º for Tm; an
empirical model based upon fitting all the DIQ data against
Tm suggests
that mirror seeing would then contribute 0.14" on the average at the extrema
in this range.
Surprisingly, we find that we are within this Tm spec 76% of the time. The
mirror is too warm 14% of the time; we could, however, easily decrease this
to 2% by a modest improvement to the mirror cooling.
To compensate for nights when the mirror is too cold, we plan to (a) replace the mirror cover with one that has gaps when open, and (b) to provide the means to extract air over the primary at night. Both of these last two improvements will be installed this summer, as will much of the ductwork for allowing us to implement improved mirror cooling in the fall. We will also be able to run the mirror cooling at all telescope positions, which will greatly increase the amount of time we're blowing cold air on the primary during the day.
Chuck Claver has continued to commission the 4-m Active Primary Support (4mAPS) this year. Wavefront analysis suggests that in typical DIQ (1.0") we should be able to shave about 0.1" from our images. Our goal is to be running on look-up maps at all three foci by 1 January 2000, and to implement a "closed-loop" system (using a wavefront camera) during the spring.
The next step we can take is to provide computer control of the f/8 tilt; this will allow us to remove the coma contribution, which become significant in parts of the sky. Further improvements may include tip-tilt at f/15, depending upon what instrumentation we plan to offer at that focus.
We are also changing a few procedures to help to improve the 4-m DIQ, primarily aimed at keeping the Mayall in better thermal harmony with the outside, nighttime temperatures. Randy Feriend and John Dunlop (NOAO/CFO) are working to understand and control the building ventilation at the 4-m. We have begun adjusting the floor cooling system to track seasonal variations better. We have also started a new policy of stowing the telescope pointing away from the zenith during the day, so that heat from the Cass cage will leak up into the dome rather than directly into the region around the primary mirror. And of course we continue to work to decrease the heat sources in the Cass cage and elsewhere in the building.
We show the most up-to-date histogram of DIQ measurements at the 4-m in the figure. Thanks to the hard work of the telescope operators, and the kindly forbearance of the observers, we have managed to obtain 1,169 measurements on 586 nights during the three-plus years of consistent seeing measurement.
Phil Massey, Tony Abraham, Bruce Bohannan,
Chuck Claver, John Glaspey, Richard Green,
Richard Wolff (the DIQ Group)