23 Nov 2007
Astronomers could soon be glimpsing farther back in time thanks to pioneering optics that are being built for the James Webb Space Telescope. Marie Freebody speaks to NASA's John Decker and Lee Feinberg to find out about the mirror and wavefront sensing technology.
When the multi-billion US dollar James Webb Space Telescope (JWST) launches in 2013, it will be a monumental day for all involved. Optical technologies that until recently did not exist will be used in space for the first time to gather infrared (IR) light from distant objects that have not yet been observed.
Among the suite of optical technologies is an innovative wavefront sensing and control (WFS&C) system and a primary mirror with a built-in support structure. But the scale of the mission means that it is rapidly approaching crunch time for researchers on the ground.
"Because many parts of the observatory take so long to build, we've actually already passed the critical design review for many elements," John Decker, deputy associate director of JWST, told OLE. "For example, the lightweight mirrors take many years of polishing and testing. We are well on our way to building those mirrors so that we finish on time to be able to integrate and test them."
Lee Feinberg, optical telescope manager at NASA, agrees. "One of the key lessons learned from other programmes, including Hubble, is to start your technology early and mature it before you get into the development stage," he said.
Once complete, there is little doubt that the finished telescope will be a unique optical instrument. "To my knowledge this is the first application of this type of complete set of technology," said Feinberg. "We have borrowed a little from other telescopes and developed many unique technologies specific to our application."
Lightweight cryogenic mirrors
The JWST will have a beryllium primary mirror, similar to the Spitzer Space Telescope. Spitzer was launched in 2003 to gather IR radiation from regions of star formation and from cooler objects in space, such as small stars. Beryllium's coefficient of thermal expansion has an extremely small variation making the telescope optics intrinsically stable to temperature variations.
About 95% of the beryllium is removed from the mirror during manufacture to make it as lightweight as possible. "The big difference is that the mirror area is about 50 times greater than that of Spitzer, but about a factor of three times lower in mass per unit area," commented Feinberg. "Since we have such a large primary mirror, we needed to make it lighter weight for launch."
The mirror is made up of 18 hexagonal segments, each measuring 1.32 m edge-to-edge. When each of the 18 individual segments is properly aligned they act as a single monolithic primary mirror. "Although they are not deformable mirrors, they do have some level of controllability," commented Feinberg. "Each mirror segment is on a hexapod with six degrees of freedom. A seventh degree of freedom comes from an actuator in the centre of the mirror that can adjust the radius of curvature."
Rigorous polishing ensures that the mirror is the correct shape. "JWST will be a 2 µm diffraction-limited telescope, whereas Spitzer was a 5 µm diffraction-limited telescope," said Feinberg. "This means that we need approximately a factor of three times better optical quality."
The mirrors will be deployed to within a couple of millimetres of the correct position and then the WFS&C system will be used to bring them into fine alignment. "The mirror will be deployed using two wings that contain three primary mirror segments," explained Feinberg. "The wings have drive mechanisms that will push them out and then a latching mechanism locks the two wings into place."
Lightweight support structure
This is the first time that a support structure will be incorporated as part of the mirror itself. It is this support structure that will be used with the WFS&C system to ensure that the 18 mirrors function as a single mirror.
"When we align the mirror using the WFS&C system, we require the mirrors to be stable for about two weeks before we go through tweaking up the mirrors," commented Feinberg. "This means that we need a support structure that is both lightweight but is also able to hold the mirrors in their aligned position when subjected to the changing temperature environment."
Both the mirror and the support structure have to be able to cool down from room temperature at launch to 30–55 K once in orbit. The semirigid mirror segment design is expected to be stable enough so that wavefront control adjustments will not be needed more than once every two weeks.
Wavefront sensing and control
The WFS&C system aligns the 18 mirror segments so that their wavefronts match, creating a diffraction-limited 6.6 m telescope rather than overlapping images from 18 individual 1.3 m telescopes.
"One of the unique features of the JWST is that we did not want to have a dedicated wavefront sensor or put edge sensors on the mirrors as this is complicated to implement," explained Feinberg. "We realized a whole wavefront sensing control capability by using images taken by the main science camera. The images can then be analysed on the ground using specially developed algorithms."
The algorithm applies an optimization method to images taken in and out of focus at different wavelengths and at multiple field points. The required mirror adjustment commands are then uplinked from the ground to the observatory.
"The WFS&C is used to align the telescope, starting from the very first moment that we get light through the telescope where it can be misaligned by several millimetres," explained Feinberg. "It goes through a multistep process, each step bringing the telescope more into alignment first with what we call a coarse alignment. We get them to within about one wavelength of light relative to each other and then finish with fine phasing."
The purpose of fine phasing is to align the mirrors to better than 25 nm. The algorithms developed for the fine phasing were first conceived to address the spherical aberration problem on Hubble.
"A phase retrieval technique was used to determine the spherical aberration on the Hubble Space Telescope by looking at defocused images and running them through a special algorithm," explained Feinberg. "The algorithm that successfully corrected these errors was then developed for the WFS&C on the JWST."
One of the biggest challenges faced by the team was to demonstrate the algorithms in a scaled model test bed of the JWST. "Once we built up the scale model, we went through the entire sequence from being a completely misaligned telescope through fine alignment and demonstrated that all of the pieces of the algorithms worked sequentially," said Feinberg.
A lot of environments had to be considered in the design and testing of the observatory. "Not only thermal testing has to be carried out, but also a great deal of mechanical testing, acoustic testing and shaker table testing," explained Feinberg. "We also analytically consider contamination and radiation environments."
It is hoped that the telescope will outlive its predicted 10-year lifetime. "We'll keep operating it as long as we have the propulsive resources and as long as it has healthy equipment operating," said Decker. "Ten years is the goal for the mission but we actually hope and believe that we will continue for a number of years after that."
Once the propulsive resources have run out, the telescope will lose the ability to maintain its orbit, which will decay over time. The observatory will then drift away in an uncontrolled orbit into space.
• For details e-mail firstname.lastname@example.org.
• This article originally appeared in the November 2007 issue of Optics & Laser Europe magazine.