RIS2 is a powerful 'add-on' to the 3.9m Anglo Australian Telescope (AAT). There is much that can be learnt about the universe from looking at longer wavelengths (the 'infrared' which we experience as heat radiation). IRIS2 allows astronomers to peer deep inside galaxies where stars are being born. Images of infrared sources taken with IRIS2 can be turned into a mask with holes cut at the location of each source. By placing this mask in front of IRIS2, together with a large prism, astronomers can then record simultaneously dozens of spectra.

IRIS2 may be regarded as significant on the following criteria.

With IRIS2, engineers and astronomers at the Anglo Australian Observatory have designed and built a world-leading instrument for astronomy at infrared wavelengths. IRIS2 is a combined imager and spectrograph for the Anglo Australian Telescope at Siding Spring Observatory, New South Wales.

IRIS2 is one of the world's first astronomical instruments able to take infrared spectra of dozens of cosmic sources simultaneously.

IRIS2 is the first astronomical instrument to use sapphire-based dispersing elements ('grisms') to create spectra. These have a higher refractive index than conventional fused-silica grisms, letting astronomers take higher-resolution spectra.

IRIS2 won the Bradfield Award for outstanding engineering achievement at the 2002 Engineering Excellence Awards and an Engineering Excellence National Award.

The multi-slit masks were designed as a collaborative effort between the Anglo-Australian Observatory and Laser Micromachining Solutions. They were made by Laser Micromachining Solutions.

IRIS2 was developed to meet a requirement for an infrared spectrograph and imager for use on the Anglo Australian Telescope. The existing instrument, IRIS, winner of a previous Engineering Excellence Award, was at the end of its useful life. Improved, larger, detectors were now available with 60 times as many pixels, and the possibility of addressing new fields of scientific inquiry would become accessible.

An instrument requirement was developed after consultation with the telescope users group. This required that IRIS2 would operate in the near infrared, 0.9 to 2.5 micrometer wavelengths, be capable of resolutions of greater than 2500 and would image a patch of sky approximately 8 arcminutes x 8 arcminutes. A Science Case, which identified potential observing programs, was prepared and a specification developed. A potential upgrade scenario was also investigated to avoid unnecessarily locking out future development of the instrument. A significant feature of the requirement was the need for a multislit capability. This would allow the simultaneous acquisition of spectra from 50 astronomical objects.

In use, IRIS2 would be mounted at the Cassegrain focus of the Anglo Australian Telescope, and it would function in a wide range of orientations as the telescope was directed at different portions of the sky. The internal deflections within the instrument had to be limited to the order of micrometers at all orientations. Between observing runs, IRIS2 would usually be removed from the telescope, so a complete system to support the instrument, monitor its 'health', and support its maintenance was required. A weight limit was imposed by the rating of the telescope instrument rotator and a space limit imposed by the size of the Cassegrain focus area of the telescope.

A series of optical designs were analysed and finally an all-refractive design was chosen. The optical components of the collimator and camera assemblies were built by Graseby Specac of the UK to the Anglo Australian Observatory design. As glass has a coefficient of thermal expansion very different from that of metals, the optical elements are mounted in housings carefully sized to accurately position the glass elements at their operating temperature, but not subjecting them to undue clamping stresses at any stage of the cool-down process. Similarly, the entire structure has been designed to be manufactured at room temperature, but achieve the critical dimensions at working temperatures. Since some components are subject to thermal gradients of 200 degrees over their length, accurate thermal analyses were required prior to final structural design.

The detector chosen for the instrument must be operated at temperatures lower than that of liquid nitrogen to minimise thermal noise, and the optical train of the instrument must be cold to minimise infrared radiation generated within the instrument. Consequently, to minimise thermal load and contamination from condensation, the instrument was evacuated and cooled by a closed cycle helium cryorefrigeration system. An unfortunate consequence of cold vacuum operation is that mechanisms are contained within a sealed vessel, neither directly visible nor readily accessible for maintenance.

To facilitate the exchange of multislit masks, which were custom manufactured for each observing field, the vacuum vessel was designed with two chambers, the smaller of which houses the slit wheel. The instrument is normally kept evacuated and cool both on and off the telescope, however to fit new multislit masks the slit-wheel vessel was independently brought to room temperature while the rest of the instrument remained cold, under vacuum. The vessel is provided with access panels, which minimised the removal of parts during a multislit mask change. Flexure suspensions were provided to keep these thin brass masks, housed within an aluminium wheel, centred on the optical axis of the instrument while preventing their thermal buckling as the assembly cools from room to operating temperature. The slits produced on the multislit masks had to be 150 micrometre wide with a tolerance of (+, -) 5 micrometres. These slits were produced by laser cutting 0.1 mm brass sheet. This technology was developed by Macquarie Research Ltd (Laser Micromachining Services).

In any vacuum system, the issue of outgassing of materials must be considered. As specified, the instrument should not require pumping while it was on the telescope, all materials within the vessel, and the vessel walls themselves, were manufactured from low outgassing materials. The vessel walls were fabricated from stainless steel. Innovative sealing techniques were required for the vessel windows, which were also optical elements.

Two helium cryo-coolers, which are mounted on vibration isolators, were used: a single stage head in the slit-wheel vessel, which has a less demanding thermal requirement, and a two stage head in the main vessel which has both a much larger cooling load, and a lower temperature requirement because it houses the detector. To reduce the cool-down time, the thermally massive contents of the main vessel were provided with a liquid nitrogen flow-through precooling system. To minimise the load on the coolers both vessels are extensively lined with multilayer insulation radiation shields.

The instrument includes a number of mechanisms: slit interchange; filter; cold stop aperture and grism wheels, and detector focus translator. The mechanisms are located entirely within the vacuum vessel. The mechanisms were designed to operate both at room temperature for commissioning and maintenance, and at cryogenic temperatures during observations, and survive the instrument cool-down process when large temperature differentials would exist and components would contract by differing amounts. Robust designs were essential.

Since the instrument normally operates in a hazardous and dark environment, remote software control of all components is provided. A typical observing procedure will involve selecting filters and grisms, pointing the telescope to the selected sky coordinates, and then making a series of observations using the Anglo Australian Observatory 2 (AAO2) controller. All of this can be done using the IRIS2 software system. Data from the AAO2 controller is processed on-line using a Sun workstation and displayed to the user in real time. The data is then recorded onto disk, along with appropriate information fully describing the observation. The data is later written to CDROM for archiving and export. Support for sequencing of observations with the instrument and telescope is provided. This allows a long sequence of observations to be written in advance. High quality graphical user interfaces (GUIs) are provided to control the instrument, detector and telescope, and to display the results of the real-time data reduction software. A data rate of 4 Mega-Bytes per second is possible. This is for the entire process, during which the data is read out of the detector, transferred to a real-time processor and then to a workstation which does the real-time reduction, display, and writes the result to disk.

The software system was largely written in-house due to the unique nature of the hardware and requirements. Where possible, external or previously written components were integrated. The software system is a networked system involving two Sun-workstations running Solaris, three VMEbus systems running the VxWorks real-time kernel and the AAO2 controller DSps. Networking hardware includes the custom fibre optic link to the AA02 controller, VMEbus backplane communications and Ethernet.

The instrument was provided with a complete suite of support equipment: off-telescope support trolley, vacuum pumping equipment, leak detector, and temperature logging and vacuum monitoring systems.

In conclusion, the design and development of astronomical instrumentation, creates the need for original thinking and often has little precedent in terms of engineering databases and design modules or mechanisms that can be reused. The organisation that is tasked to conduct such projects therefore has to develop a clear picture of the science need and progress using a process that allows review and critique of the developing design and operating vision.

In order to meet the project requirements and balance the conflicting organisational demands for expert and specialist resources across competing project requirements within the AAO, a matrix form of organisation was established. This structure aimed to balance internal resource demands to bring the appropriate level of skill and experience to bear on the IRIS2 project and ensure the planned outcome was achieved.

The key staff from the AAO who were involved in the IRIS2 project and their roles, were:

Project Engineer: Roger Haynes. He controlled all technical aspects of the project and liased between the three functional groups.


Instrument Scientist: Peter Gillingham. Optical design executed in tandem with Prime Optics, an external consultant.


Mechanical Design Engineer: Greg Smith. Smith did most of the early mechanical concept design work. Later he did some of the detail design work specifically: detector translator; grism wheel assembly, collimator, multi-slit mask holder, worm assembly flexure design; optical baffling; structural design; and pupil imager.


Mechanical and Cryogenic Engineer: Vlad Churilov. Churilov worked on thermal design and vacuum system design including vessels, sealing including vacuum tight lens mounts, material choice, vacuum system integration, as well as the detector mount and all driving mechanisms.


Resource Centre Manager (Mechanical): John Dawson. Overall coordination of mechanical group effort. Subcontracting of work. Checking of detailed mechanical designs.


Mechanical Engineer and Designer: Mark Hilliard. Detailed mechanical design as required by the lead designers.


Electronics Engineer: Brian Hingley: Design and testing of electronics for instrument mechanism control.


Resource Centre Manager (Electronics): Lew Waller. Overall coordination of electronic group effort. Design, development, and testing of AAO2 Controller digital electronics and firmware.


Electronics Engineer: John Barton. Design, development, and testing of AAO2 analog electronics. Testing of the Rockwell Hawaii Infrared detector.


Resource Centre Managers (software): Keith Shortridge and Tony Farrell. Develop and maintain software estimates and schedules. Assign software resources and resolve software related planning conflicts.


Software Engineers: Jeremy Bailey; Keith Shortridge; Tony Farrell; John Straede; Anthony Dunk; Serge Ivanoff; Niki Frampton; Peter Innes. Software design and development, testing and commissioning. Support for mechanical and electronic components.


Project Management: Chris Evans. The principal responsibility was to deliver the project within budget and time limitations in accordance with the technical specifications and with objectives set by AAO management.


Task Manager: Gabriella Frost. Managing the various tasks of the IRIS2 project at the integration and testing stages of the project.

The multi-slit masks were used in the IRIS2 at the Anglo-Australian Observatory.

The Anglo Australian Observatory (AAO) until recent years had developed instruments only for its own telescope, the Anglo Australian Telescope (AAT). One of these was IRIS, the forerunner of IRIS2. During the first half of the 1990s, the AAO developed an instrument called 2dF, a previous entrant for the Engineering Excellence Awards, which included two multi-object spectrographs. This instrument, which went into full service observing in late 1997, earned the Observatory a high reputation in optical astronomical instrument building. The AAO won a contract to supply the European Southern Observatory (ESO) VLT telescope situated in Chile with a conceptually similar instrument called OzPoz, which is a robotic device that positions optical fibres to take the light from distant stars and galaxies and spreads it into their component colours. At this time, the AAO took a strategic decision to implement a more rigorous project management system. During the latter part of 1998 and into 1999, it went through a steep learning curve trying to apply good management practices on its developing instruments IRIS2 and OzPoz.

IRIS2, as a cryogenic instrument, had many technical risks associated with its design and development. Consequently, the eventual cost of the instrument was significantly more than was predicted at this time. Nevertheless, many other observatories around the globe have also attempted infrared instruments and have ended up significantly over cost and time budgets, as well as delivering instruments that have proven unreliable in service. One example of a technical problem that resulted in major time delays and cost increases for IRIS2 was the stepper motor operation in the cryogenic environment. Although the contract specified that the stepper motors, which were designed and made in Britain, were to work at liquid nitrogen temperatures, they failed to do so. Consequently, the AAO engineers and technicians designed and manufactured modified stepper motors and these have shown to work effectively in the cryogenic environment.

IRIS2 multi-slit masks (2), infrared spectrograph and imager components, metal, used by the Anglo-Australian Observatory, designed by Laser Micromachining Solutions / Anglo-Australian Observatory, made by Laser Micromachining Solutions, Sydney, New South Wales, Australia, 2002

Multi-slit masks for the IRIS2. Square black plates with an array of rectangular slits and company and University logos cut out of the metal.

Designed: Anglo-Australian Observatory; Sydney; 2002

Designed: Laser Micromaching Solutions; Sydney; 2002

Made: Laser Micromaching Solutions; Sydney


Used: Anglo-Australian Observatory; Sydney

Used: Anglo-Australian Observatory; Coonabarabran, New South Wales

See parts for marks information

2004/124/4

2002

80 mm

70 mm