THE UNITED KINGDOM INFRARED TELESCOPE
ANNUAL REPORT
1998

4. Approved Programme

4.1. Completion of the UKIRT Upgrades Programme

At the beginning of 1998 the Upgrades Programme was formally complete except for insulation of the dome floor and the primary mirror cooling system. During the year the benefits of the upgrades were dramatically demonstrated by the results of a seeing monitoring programme which confirmed that UKIRT's images now only rarely approach anything as bad as 1 arcsec FWHM; users now tend to regard 0.7" as ``poor'' seeing.

4.1.1. Telescope Optical Performance

Between 13 February and 29 September 1998 a standard protocol was used to estimate the delivered K-band image FWHM in a uniform way, ideally twice per night, whenever the imager IRCAM3 was in use. A total of 143 measurements were secured, 41 of them in September, when IRCAM was continuously on the telescope.

About 75% of the images in the resulting data-set were taken at the so-called x 1 pixel scale (0.28"/pixel) and about 25% with the x 2 magnifier (0.14"/pixel). The two image size distributions differed notably on account of undersampling of the images with the larger pixel scale, and sets of consecutive measures with the two scales were acquired to construct a correction curve. This was used to correct each observed image FWHM to a ``fully-sampled'' equivalent. The resulting corrected distribution was remarkable in that a significant fraction of the FWHM measurements appeared to be within 0.05" of that of a diffraction limited image (nominally 0.108" at 2.2 µm with UKIRT's aperture and central obstruction).

It was not initially believed that UKIRT really was regularly offering images near the diffraction limit, especially as this conclusion depended on a correction for sampling effects by a factor of more than two. However, on 8 September 1999 Antonio Chrysostomou and Thor Wold were able to secure consecutive well-focussed images with no magnifier ( x 1) and with the x 2 and x 5 magnifiers. The latter, yielding pixels only 0.057" square, could fully sample images quite close to the K-band diffraction limit. The results not only confirmed the undersampling correction curve, but yielded images of extraordinary quality. The mean of the FWHM of three 20 second exposures was 0.178"; the best image had formal FWHM =0.171" with a Strehl ratio of about 0.25 (i.e. the peak intensity in the image is 25% of that in a perfect image sampled with the same pixel distribution). We believe that these may still be the best (sharpest) images yet secured by a ground-based telescope without the use of high-order Adaptive Optics (i.e. higher than tip-tilt). These images are all the more remarkable because the TSS considered that the night, while good, was not of the very best. This was in fact consistent with the histogram of corrected images: the images with the x 5 magnifier fall in the next to smallest populated bin; at least 10% of the sample images appear to be better than these.

Over the period of the seeing measurement campaign the median corrected image FWHM was 0.433". 90% of image FWHMa were below 0.86" and 10% were less than 0.15", i.e. within 0.05" of the diffraction limit. Over the month of September the median corrected image FWHM was 0.265" and the worst image had FWHM 0.563".

The seeing campaign was terminated when the new imager UFTI had its first light on 30 September 1999. While the sampled period in 1998 may have been of above-average seeing overall, UFTI, with its 0.091" pixels, has since yielded many images with FWHM between 0.3" and 0.4" and a few with FWHM < 0.3".

4.1.2. Dome Ventilation System

The main Dome Ventilation Sytem (DVS) was in general use by the spring of 1998, when 13 of its 16 apertures were fully functional. The remaining three require heavy mechanical (crane) work to correct defects, and at the time of writing are not yet operational. Nevertheless, as indicated in last year's report, the average inside-outside temperature differences are now close to zero for a significant fraction of the time (see below). The DVS has proved reliable in general use but is vulnerable to jamming if ice forms on the guides of the roller doors, and quite careful procedures must be followed to avoid this.

In mid-1998, after a review of the dome thermal behaviour by a Co-op student, Patti Smith, fans were installed on the South column platform. These were pointed upwards so as to circulate air through the top of the dome, to prevent heated air accumulating there and heating the top-end steelwork. The plant room ventilation system was set to extract air (at around 10 volumes/hour) from the dome, which it does at floor level, thereby removing any warmed air moving down from the upper parts of the dome. The measured circulation rates and the lack of any temperature rise in the lower structures indicated that this strategy worked well. For the experiment, during the month of September, the fans were switched on every second day and the averaged results from ``fans on'' and ``fans off'' days compared. The fans reduced the warming of the top-end by about a third and reduced the average inside-outside temperature difference from about half a degree to approximately zero for more than half the night. On the nights with the fans on the image FWHM averaged 0.04" smaller than when the fans were off, a 15% improvement in resolution. Further experiments with increased fan power have been conducted in 1999.

4.1.3. Dome Floor Insulation

As outlined in the previous report, safety considerations led to the abandoning of plans to insulate the upper surface of the concrete dome floor with removable foam-plywood panels. Instead, the underside of the floor was insulated to minimise tranfer into the concrete, and hence the dome air, of heat from the crew rooms, etc., below. At the same time changes were made to the downstairs ceilings and doors to improve the confinement of heat to the areas in which it is needed. The design and management of the project were largely carried through by Patti Smith. The installation was completed early in 1999.

4.1.4. Primary Mirror Cooling

Further progress was made with the installation of the cooling system as effort allowed. By the end of 1998 detailed design was complete and installation of coolant plumbing was well advanced. This continued into 1999.

4.1.5. Secondary Mirror Replacement

As noted last year, the new secondary mirror displayed several defects, principally a turned-down edge, which reduced the limiting achievable Strehl ratio to around 70%; print-through of the light-weighting pattern, which makes the use of Adaptive Optics difficult or impossible; and trefoil aberration induced by thermal stress at the mounting points, further reducing the limiting Strehl ratio and, with the light-weighting pattern, generating a hexagonal pedestal under the images. In 1997 an agreement was accordingly reached with the MPIA for the provision of a new secondary mirror without these defects.

In early 1998 a careful review of potential suppliers was undertaken to establish technology availability, risks and likely costs. This established that obtaining a new mirror from a single supplier in the USA would certainly be expensive, probably between $150 k and $200 k, i.e. at least twice what the MPIA estimated their managed procurement would cost, and therefore involving significant costs (50% or more of the total) for PPARC. After careful consideration of risks and costs the MPIA were authorised to proceed with the procurement as proposed. Fabrication of the slightly oversized optic, testing through the rear surface, followed by reduction to the specified diameter to remove residual turned-down edge (TDE) on the optical surface, was carried out by Horst Kaufmann Prazisionsoptik, Crailsheim-Wittau, Germany, in late 1998. In early 1999 the mirror was lightweighted by grinding at BNM GmbH, Jena, Germany, and the back surface was then acid-etched for stress relief by Carl Zeiss, Oberkochen, to eliminate print-through of the lighweighting pattern, which is prominent in the original mirror. Finally, new athermal mounts were glued into the appropriate pockets in the back of the mirror. These were designed by Physik Instrumente, Waldbronn, Germany, and R.J.Bennett at the UKATC, to eliminate the thermal stresses which at normal operating temperature impose significant R⁵ and R³ trefoil aberrations on the original secondary. ¹