Telescope Structure and Optics in Cassegrain Mode

Telescope Structure and Optics in Cassegrain Mode

Top-end and Secondary Mirror systems
Primary Mirror and Active Figure Control (this describes both Ciassegrain and wide-field mode)
Dichroic Tertiary Mirror
NB: See also the bottom-end systems which control the tip-tilt secondary and its positioner. 


UKIRT’s primary mirror is about 1/3 as massive as was normal at the time of its construction in the mid-1970s. A key element of its design is therefore the mirror cell, which is very massive and rigid, with webs over a metre deep. The cell is suspended on Serrurier trusses from a midsection which carries the declination axis, and which in turn supports the topend via a second set of trusses. An outline drawing of the telescope structure can be viewed here. 

The rectangular centre section is of unusual design: two substantial beams to east and west (the “Dec Beams”) are linked by two thin tension members (the “tie beams”) on their north and south ends. Tension in the tie beams is maintained by tension in the declination axis, which is maintained at ~20 tonnes by two inflated stainless-steel toroids located in the yoke. 

The thin tie beams were intended to permit use of a coude focus with minimum obstruction in those attitudes when the central structure must needs obstruct the beam; in the event the coude focus was never used for observing, but the structure of the center section is exceptionally light and efficient. A view of a solid model of UKIRT in its building, showing this structure, is available here.

The telescope is equipped with a number of temperature probes which allow both steelwork and adjacent air temperatures to be measured. These are used for diagnostic and for control purposes. In particular, the temperatures of the steelwork are used together with a model of the telescope’s elastic behaviour to predict changes in the telescope’s length are fed back as corrections to the hexapod positioner so as to keep the instrument in focus at all temperatures and attitudes. 

Top end and Secondary Mirror systems

Topend structure 

The telescope top-end, along with the secondary mirror, its tip-tilt actuation system, its precision positioning system and the associated bottom-end fast guider systems, was supplied by the Max Planck Institüt für Astronomie (MPIA) in Heidelberg. The top-end comprises a substantial tubular steel ring which bolts to the fixed top-end ring above the upper Serrurier trusses of the permanent telescope structure. It carries four vanes which rise as they converge on the central boss. The vanes are of 16mm stainless steel and are surfaced with constrained-layer damping to reduce the transmission of vibrations from the fast “tip-tilt” secondary mirror to the telescope structure and eventually the fast guider sensor, which could cause ctastrophic resonances. 

The converging vanes meet the square central boss, to which they are bolted, tangentially; consequently the vanes are not oriented exactly in the cardinal directions and diffraction spikes should not be used for accurate image orientation. 

Secondary mirror positioning system: the Hexapod 

The secondary mirror and its actuation system are in turn supported by a hexapod precision positioning system manufactured by Physik Instrumente of Waldbronn, Germany (PI). The six legs, containing precision screws driven by DC motors, are extremely stiff and can position the secondary mirror with an accuracy of ~2 microns in translation and a few arcsec in rotation. 

Secondary Mirror 

The original tip-tilt secondary mirror was delivered with the new top-end systems in August 1996; these (see below) produced an immediate improvement in image quality and telescope performance by the elimination of image blurring due to telescope vibrations. However the lightweighted mirror suffered a turned down edge, significant print-through of the lightweighting pattern (dramatically apparent in the out-of-focus image, right), r3 and r5 trefoil, and some spherical, aberrations caused by thermal distortion due to the mounting system. Overall the defects of the secondary were believed to reduce the limiting deliverable Strehl ratio of UKIRT’s images (all else being perfectly adjusted) to around 50%.

Accordingly, in 1998-99 the MPIA procured a new secondary. This shows none of the above defects (see out-of focus image, left). Like its predecessor, the new mirror is made of Schott Zerodur low-expansion vitreous ceramic, selected for very high optical uniformity, and was also figured by Präsizion Optik of Crailsheim, Germany, by testing through the back. However this mirror was made ~14mm oversize, so that the edge could then be ground off to eliminate turn-down. Again, like its predecessor, the figured mirror was then 60% lightweighted, in this case by diamond grinding the hexagonal pattern into the back. This was carried out at ZBNM Technisches Institüt, Jena, Germany. In the other critical departure from the previous process it was then stress-relieved by acid-etching at Karl Zeiss, Oberkochen, Germany. At the MPIA the mirror was fitted with three re-designed athermal mounts, fabricated of Invar and Zerodur by PI, before being shipped to the telescope. It was installed on 14 June 1999.

Summary description of the new secondary mirror

The following lists the specifications for the new secondary mirror. The only difference between the old secondary mirror (K1) and the new secondary mirror (K2) is the central hole size. For K1 this was 85.0 mm. 

Part of MirrorSpecification
Clear aperture313.4 +/- 0.1 mm
Central Hole92.0 +0.2 -0.0 mm
Radius of curvature1723 +/- 2 mm
Conic constant-1.326 +/- 0.005
Surface quality30-50 scratch and dig
Surface figure:
- Image locations: - file:///ukirt/export/mauiola/www/html/UKIRT/telescope/oldsecoof.jpg– over entire diameter< 110nm P-V
Over any 100mm diameter region< 70nm P-V
Over any 50mm diameter region< 35nm P-V

Tip-Tilt system

The secondary mirror is equipped with a high-bandwidth articulation system designed and manufactured by PI. The mirror is attached to three piezo-electric actuators and can be rapidly tilted in any direction to move the image by up to ±17″ on the sky. It is normally driven in antiphase with a similarly-mounted momentum-compensating mass which corrects for 90% to 95% of the disturbance torques generated, and has a bandwidth well in excess of 100 HZ. Its principal function is fast “tip-tilt” image stabilisation, correcting image movements of order an arcsec caused by telescope mechanical vibrations (e.g. “wind shake”), drive errors and atmospheric seeing-induced image movement. The tip-tilt system is controlled by a Fast Guider at a telescope focus, which measures the position or the image of a guide star up to 100 times per second and sends corrections to the secondary control system to stabilise it. This system can also be used for chopping between two points on the sky up to 20″ apart. (See Acquisition and guiding systems for more details of the Bottom-End systems which control the secondary’s actuation and positioning systems.) 

Primary Mirror and Active Figure Control in both Cassegrain and Wide-Field Mode

The thin primary mirror is relatively flexible, and consequently amenable to control of its large-scale optical figure by quite modest forces. The Table summarises its properties. 

Mirror PropertyValue
MaterialOwens-Illinois “CERVIT”
Outer Diameter3802 mm
Inner Diameter1028 mm
Outer edge thickness287 mm
Inner edge thickness192 mm
Mirror Mass6425 kg

The components of the support system are shown schematically in this figure.

The mirror is supported axially by 80 pneumatic “bellofram” frictionless pistons, and radially by 24 lever arms glued to the sides. The belloframs can be controlled as one (the original arrangement) or as three separate sectors (the default arrangement since the support system was upgraded in 1994). In addition there are three radial positioners (located tangentially to the edge), three axial positioners equipped with load cells and controlling the sectors of belloframs by a servo loop which adjusts their load to zero. 

Around the edge of the mirror are 12 force actuators. These pneumatic devices can apply up to ±500 N with a resolution of 1 N in an axial direction. By this means the primary can be bent to correct the following low-order aberrations: 

  • astigmatism up to ±6.9 microns 
  • triangular coma (trefoil) up to ±1.8 microns 
  • quatrefoil and cinquefoil up to ±0.5 and ±0.2 microns respectively 
  • spherical aberration up to ±0.5 microns 

… where these limits are P-V wavefront errors.

The last-mentioned aberration is corrected by generating a substantial change in overall curvature. This is an r2 function, but the process produces some aliasing into r4 spherical aberration along with considerable defocus from the r2 effects. The latter is easily corrected by refocussing. 

The amplitudes listed are those achievable if the entire available authority of the actuators is devoted to the single aberration; in practice several aberrations must be corrected at once and at the time of writing residual some spherical aberration (about 180 nm RMS on the wavefront) cannot be completely corrected. 

Dichroic Tertiary Mirror 

All instruments on UKIRT are mounted at broken Cassegrain foci facing one of four exit ports of the Instrument Support Unit (ISU), attached to the bottom of the mirror cell. The ISU contains a the rotating tertiary mirror which normally carries an IR-reflecting dichroic coating, which directs the IR beam to the to the selected port, while allowing visible light to continue downwards to the acquisition and guiding systems. Fine adjustment of the tertiary mirror in rotation and in tilt allows accurate alignment of the instrument acceptance cone with the secondary mirror. (Alternatively put, it can be adjusted so that the delivered f/36.4 IR beam emerges emerges horizontally from the port.) 

Two tertiary substrates are of BK7 glass figured to a high optical quality, and a spare of selected float glass is available for special-purpose coatings if need be. No optical defects originating in the tertiary mirror have ever been identified. 

The normal dichroic coating was designed by Carl Zeiss (Oberkochen). It is a proprietary multilayer silver-dielectric combination “formula 2”. When in good condition (it is quite fragile) its properties are as follows. 

  • Transmission ~50% (TBC) of incident “CCD visible” light. 
  • Reflection: >50% longwards of ~850 nm, >97% longwards of 1 micron. 
  • Emissivity: Probably ~2% longwards of 2 microns. 

A further float-glass substrate with a gold-on-chromium coating is also available at the telescope.

Back to top of page