Space telescopes provide a good example of the advantages of an integrated program, with its vastly greater capabilities for deployment and operation, in particular the large object vehicle and ion propulsion for orbit transfer.
With the large object vehicle a 10 meter mirror can be put into LEO in a single launch, with no sectioning or deployment required. Adjustments can be made at the space station, and the telescope moved to any desired orbit, at a weight as high as 20 tons. The issue of placement should be re-evaluated within the context of an integrated program and its greater capabilities, and other parameters also.
Design problems include the following. The mirror must maintain its shape, with a structure whose weight is moderate, 20 tons at most. The mirror must withstand the launch well enough to be calibrated. The temperature and regulation strategy must be determined. The material and thickness of the bearer of the optical surface, and its method of attachment to the truss, must be decided on. The telescope must retain its optical shape during an observation, with any active controls in operation. The power consumption and method of supply must be determined; for one possibility, a standard power module could supply power to a suite of telescopes via microwave transmission. Although other possibilities can be implemented in an integrated program, data from the focal plane is generally transmitted electronically. Channels and archiving methods of sufficient capacity must be devised.
There seem to be two overall strategies, a "big dumb truss", which withstands the launch and has little active control, and a lightweight "smart" truss, which might use more active control methods, a thin mirror, etc. A major consideration in design decisions should be the marginal cost of additional units. In the case of a lightweight truss, calibration is supposed to be feasible after unfolding for very lightweight mirrors. For another approach, during launch the light truss can be supported along members by contact with resting surfaces on a heavier truss. In orbit the light truss is separated from the heavy one; the heavy truss is returned to earth for reuse.
The value of 10 meters mentioned above was used for illustration; the diameter of a basic space telescope might be between 8.5 meters (the largest mirrors currently made) and 13 meters (the capacity of the large object vehicle). An interesting possibility is to make mirrors of various sizes as arrays of hexagonal mirrors of outer diameter 5 meters. A seven mirror array then fits in the large object vehicle. The atomic unit might be a tunable thin mirror attached to a truss. Trusses are assembled into a "supertruss"; this is tunable and also can have disturbance rejection (e.g.,after steering). The supertruss tuning need only get the mirror surfaces within the limits of their tuning range. Alternative methods involving secondary mirror tuning can also be considered. The equilibration time needs to be somewhat less than a typical observation time. Steering might be done with jets, since in an integrated program replenishment is provided. A low altitude might be preferred due to operational considerations.
Quite large telescopes might be possible with these methods, say 6 or 7 rings (126 or 168 mirrors). A factory with adequate production rate is needed. Spinning might be possible for off-axis mirrors by spinning them in pairs at the ends of a boom. A variable force grinder then grinds a mirror to a shape adequate for polishing. The cost of such a telescope might be kept under $3 billion.
Interferometry with multiple 10 meter (or larger) telescopes is of interest. The telescopes must have the ability to send a beam to a combiner box; this can do all the work, so that minimal additional facilities are required on the telescope (and hence can be built in from the start). The integrated program might relax design constraints on telescopes at other wavelengths, or on multi-wavelength configurations. Radio telescopes are of interest, in particular for interferometry with a baseline equal to the diameter of Earth's orbit.
Various of the foregoing suggestions are already under consideration, for example the use of lasers in coordinating space telescopes for interferometry, and various designs for orbiting mirrors. The new ingredients are the large object vehicle, and ion propulsion for orbit transfer; larger and heavier objects can be put in LEO and moved to any orbit, even solar. Costs of large space telescopes might be reduced to the point where earthbound telescopes might even be obsolete (although this is a complicated issue).