Still in its early stages of design, NASA's Next Generation Space Telescope (NGST) is slated to launch in 2009. All three optical concepts now being considered for the NGST are based on the same approach: an 8-m primary mirror formed of rigid segments that unfurl like the petals of a flower. Although the segments themselves will be lightweight (made ribbed and hollow), the total payload mass will be on the order of 3000 kg, requiring a large, expensive launch vehicle such as an Atlas IIAS-class rocket.
There may be other ways to make a large space telescope, however. The dream of an 8-m space telescope that is much lighter and less costly to launch has led Aden and Marjorie Meinel, two researchers retired from the Jet Propulsion Laboratory (JPL; Pasadena, CA), to re-examine an old idea—the inflatable membrane mirror—with encouraging results.
Such a mirror is made of two membranes attached to a circular tensioning ring. When the enclosed space is pressurized, the membranes expand into bowl shapes. If one membrane is aluminized and the other transparent, a focusing optical surface results. Unfortunately, the shape taken by a pressurized membrane of uniform thickness is an oblate spheroid that produces many hundreds of waves of spherical aberration, making it useless for telescopy.
In an attempt to solve this problem, the Meinels have developed a theory describing the shape of a pressurized membrane having a radially varying thickness—a concept suggested by James Breckinridge of the JPL. In particular, they calculated that if the membrane thickness varies quadratically as a function of radius, than the inflated membrane itself takes up a paraboloidal shape. (The Meinels have collaborated for many years on the design of optics; previous to their stay at the JPL, they were both at the Optical Sciences Center at the University of Arizona. Aden Meinel was the founding director of the center.)
A membrane mirror could take one of two forms: flat when unpressurized or preshaped into a paraboloid. Although both take a paraboloidal form when pressurized, a preshaped mirror can potentially be made deeper, while a flat mirror may be simpler to fabricate. Methods to make preshaped paraboloids take advantage of the fact that when a liquid-filled container is spun under the influence of gravity the fluid surface takes a paraboloidal shape—a technique ideal for casting the membrane itself, as well as its underlying mold. In one example, a 1-m-diameter mold is cast at a rotation rate of 2.9770 s/revolution. Changing the rate to 2.9807 s/revolution, the mirror itself is cast. The result is an f/1.1 paraboloidal membrane; if its central thickness is made to be 0.100 mm, the resulting edge thickness is 0.143 mm. When pressurized, the membrane remains paraboloidal.
The researchers are examining a configuration for a space telescope that is almost entirely inflatable. A cylinder of inflated rings tensions the primary mirror and, with the help of a carbon-fiber truss, positions the secondary. By itself, a membrane mirror with radially varying thickness would produce a wavefront with a residual error in the 10-100-wave range. With the aid of a wavefront corrector, an optical-passband telescope including such a mirror could potentially operate in the 0.5-12-µm region at a final error of 0.1 wave rms for at least part of this range.
There are many hindrances to this concept, the researchers point out. The optical surface for a first-surface membrane mirror will not be the one in contact with the mold, thus lowering surface quality. A membrane mirror will have to be folded into a compact package for transport into space; no good two-dimensional folding technique yet exists. Upon pressurization, the mirror will stretch, possibly cracking the coating. There is not yet a good design for an inflatable tensioning ring, or for the points of attachment to the mirror. But the lure of cheaply launched large space telescopes may yet spur scientists to tackle these problems, although not in time for the NGST.