Left1

| Add RSS Feed

INDUSTRIAL LASERS: Carbon-fiber spectrographic masks are precision laser cut

Continuous-wave Nd:YAG-laser-based cutting system will supply plates for multislit spectroscopy for Gemini North and South telescopes.

John C. Hamilton

The Gemini 8.1 Meter Telescope Project, a joint international partnership of the United States, United Kingdom, Canada, Chile, Australia, Brazil, and Argentina, will ultimately have two operational 8.1-m telescopes

The telescopes stand out from their contemporaries in several ways. Both have identical 8.1-m mirrors that will explore the skies over the northern and southern hemispheres, collecting optical or infrared radiation from space. They should allow astronomers to observe the universe at infrared and visible wavelengths with more clarity than has ever been possible from the Earth's surface.


FIGURE 1. Continuous-wave Nd:YAG laser cutting machine will slew up to 1 m/s (travel through air between cuts), yet stop and cut slits in spectroscopic masks to an accuracy of 0.25 µm

A key piece of equipment required for each telescope is a Gemini multiple-object spectrograph (GMOS), which was developed in a collaborative project between teams in the United Kingdom and Canada. The device will provide 0.36

For multislit spectroscopy, observers will use the GMOS imaging mode to record an image of the sky and then select the sizes and locations of slits to be cut in a mask (see Fig. 1).1 Several masks may be necessary for a single image field. With its field of view, the GMOS will be able to locate several hundred slits in a single mask, which will sit at the focal plane in the GMOS optical train (see Fig. 2).

To satisfy targeted specifications for spectral resolution, velocity accuracy, and sky subtraction, masks will require precision fabrication. The width of a slit will be much narrower than on plates cut for other telescopes, and the slit opening will need to be of uniform width and free of jagged edges. The fabrication system will also have to meet tight positioning tolerances for each slit.2

Narrowing the fabrication field

Although laser cutting was ultimately chosen for mask cutting, the Gemini project team also explored other precision fabrication options, including photochemical etching, water-jet cutting, electrical-discharge machining, and computer-numerical-control machining. Researchers examined processes for their capabilities to cut to tolerance in a variety of work materials, especially the carbon-fiber sheet ultimately chosen for use in Gemini masks.

Although aluminum is a common mask material for multiple-object spectrographs, its high coefficient of thermal expansion precluded its use in Gemini masks. Based on the possible temperature variation during GMOS operation and the influence this could have on slit-to-slit positioning, the researchers only evaluated work materials with a coefficient of thermal expansion lower than that of stainless steel and with commercially available thicknesses of less than 1 mm. The material selection also depended on the mechanical stiffness of each. Maximum sag could not exceed 1 mm in the large masks required for Gemini. Carbon-fiber sheet outperformed metals in both areas.

The carbon-fiber material selection narrowed the list of fabrication candidates because electrical-discharge machining can only cut electrically conductive materials. Photochemical etching was also eliminated because current equipment reportedly could not economically meet the precision and automation requirements. Laser cutting outperformed the other processes in areas related to slit dimensional and edge tolerances, as well as cutting speed, feed, and flexibility.

The laser mask cutter selected by the Gemini project team is a Model 104-C machine from Advanced Recording Technologies (Escondido, CA) with a continuous-wave Nd:YAG laser and high-speed linear motor stages. The device can cut precision slits into 10-in.-square masks that can be replaced every hour or so with new slit patterns for observations in different regions of the sky.

According to Advanced Recording Technologies, which plans to develop mask cutters for other telescope projects, some models can cut slits within 0.1 µm of accuracy and process masks that are 100 mm to 1 µm in diameter and 0.5 mm thick. Up to 2000 slits can be cut per hour in stainless steel, tantalum, epoxy/carbon, or anodized aluminum. Normal average-power requirement for 5-mil stainless steel is 10 W at 1.06 µm.

Mask cutter goes on-line

The mask cutter for both Gemini North and South now resides at the Gemini base facility in Hilo, HI. During preliminary experiments with laser cutting, the researchers found that such equipment can cut slits with edge straightness and smoothness within the 1.2-µm functional requirement. In addition, they estimated that a 200-slit mask could be cut in about an hour, based on cutting a 120-µm x 3-mm slit in two passes of four seconds. Actual processing time

For budgetary and other reasons, systems associates running the machine improvised certain additions. For example, Gemini project researcher Jim Stilburn of the Herzberg Institute of Astrophysics (Victoria, BC, Canada) developed the system's autofocus unit, which fits between the mask and the bottom optics of the laser. The 10 x 10-in. masks are held to the worktable by a frame around their edges, which means there could be a lot of play in the system. The autofocus device relies on a slight vacuum suction to automatically pull the mask up to its adjustable collar. After focusing the laser, the system operator can then move the laser across the mask without worrying about ripples or small variations in its surface or even tilting of the xy positioning stage

In addition to relying on autofocusing to reduce errors, machine operators calibrate the system with mask-to-detector mapping. They laser cut a matrix pinhole mask that is then placed at the GMOS focal plane. The pinholes are imaged by the detectors and their positions checked against the theoretical xy-stage positions. According to researchers involved in developing the functional design for the mask-cutting process, the discrepancies form a fitting surface that is the transformation between their position on the detector and the corresponding xy-stage position requirements. Mapping ensures that alignment of the slits to objects is as accurate as the limit of the stage's random error.

To adjust laser power before cutting masks, it is also necessary to have the shutter open and the laser running at full power, which means there must be some work material to absorb the heat. While it is common to use ceramic bricks for this purpose, they were in shorter supply than lava rocks, which machine operators report are low cost and work quite well.

Targeting return on investment

To track masks, researchers also are creating a dedicated in-house database that will contain the initial field-image exposure and the mask-definition files by name. For each definition file, a blank mask will be laser cut and then numbered with a barcoded label for tracking purposes. After successful observations have been obtained with the mask, it will be retired to the main archive library for a period of up to two years. The mask cutter will also produce various calibration masks for storage in this collection.

Researchers anticipate that the laser cutter will begin mask production for the GMOS in about a year. Currently, they are fine-tuning the process with cutting jobs for partner countries in the GMOS project. For example, while cutting masks for Australia's Anglo-Australian Observatory, engineers identified long-term durability problems with the current carbon-fiber mask material. Because carbon fibers in the three layers of the material run in the same direction, fracture lines can develop along the grain when holes are cut. Stilburn is now talking to materials suppliers about the possibility of using a carbon-fiber material with the grain in the middle layer running perpendicular to that in the other layers, which should improve material rigidity and thus mask life.

REFERENCES

  1. K. Szeto et al., "Fabrication of narrow-slit masks for the Gemini multi-object spectrograph," Gemini Project; gemini.edu (1997).
  2. T. J. Davidge, R. Murowinski, and J. Allington-Smit, "GMOS Functional Requirement," Gemini Project; gemini.edu (1996).

Masks enable multislit spectroscopic observation

Observers using a Gemini telescope receive the image of a specific sky field obtained at an earlier observing session and select the astronomical objects, such as planets, stars, portions of nebulae, or galaxies, in the field for which he or she wishes to collect spectroscopic data. This information determines the slit pattern to be cut on a Gemini multiple-object spectrograph (GMOS) mask. The photo illustrates examples of a close galaxy selected by an observer using Gemini's Canada-France-Hawaii telescope (Mauna Kea, HI).

As in most spectrographs, with the GMOS the light first passes through a slit and is collimated before it is dispersed in the spectral dimension by the grating or prism (or even grism). If one passes monochromatic light through the system, an image of the slit is seen. So the precise size and quality of the slits are critical to the quality of the resultant spectra.

A precision-cut mask, located at the focal plane of the GMOS, take the place of the conventional slit. If only one slit is cut on the mask, then this mode of observation is identical to classical spectroscopy. What makes this such an advance is the multiplexing of observation time made possible by allowing many slits to be uniquely positioned to permit observation and analysis of many celestial objects simultaneously. Thus, an entire field can be observed in one exposure, rather than many

The GMOS masks not only allow light from the desired objects to pass through, but they also block the light from any and all other objects in the field. This permits capture of individual spectra of adjacent objects without overlap and confusion of sources, which was common in objective-prism spectroscopy work. In that technique, a prism was placed in front of the telescope and the field imaged. The observer received spectra of everything in the field, although overlaps and differences in individual source brightnesses rendered some of the spectra useless. Also, this technique was used only for low-dispersion work, suitable for surveys. The GMOS masks allow for high-dispersion spectra, and one can select sources of similar brightness for different masks of the same field to allow suitable exposure times.

JOHN C. HAMILTON is a systems support associate with the Gemini Observatory in University Park, 670 North A'ohoku Place, Hilo, HI 96720; e-mail: jhamilton@gemini.edu.

Fri Oct 01 00:00:00 CDT 1999


| Add RSS Feed

Laser Focus World Article Categories:


Search Products Buyer's Guide >

Search Industry Specs >

Search Industry Jobs >

Magazine & E-Newsletter Subscriptions >

ADVERTISEMENT
Middle
ADVERTISEMENT
Right1
Right
Right2
Right3
Bottom1
Bottom2
Bottom3