Geoff Andersen, Ph. D.

Laser and Optics Research Center,

USAF Academy, CO 80840

The Starshine mirrors are fabricated from polished aluminum metal, which is given a protective coating of silicon dioxide. On visual inspection, many of these mirrors still have slight scratches present which will tend to scatter some of the reflected light into large angles, which may reduce the brightness of the solar reflection to a ground observer. In order to assess just how severe this scattering is, we used the set-up shown in Fig. 1. A laser (with a wavelength of 532nm) was reflected off the surface of a test mirror, and a horizontal scan of the beam was made 1.6m downstream.

The results are shown in Fig. 2, with a "perfect" optical reference mirror shown for comparison. Three Starshine mirrors were chosen at random, and configured in a worst case scenario, with the axis of greatest scatter oriented along the direction of the scan. At first glance, it would seem that there is a very large amount of scatter at large angles, compared to a conventional optical mirror, but a closer look at the amount of power contained in these "wings" shows this isn’t so. In fact, by calculating the area under the curves, it is found that 99.9% of the overall flux is contained in the region where the two curves match very closely (from 11mm and up), and the scattered component outside this region represents only 0.1% of the overall reflected light. This means that the Starshine mirrors will reflect nearly all the incident light the same as a high-quality (expensive) glass mirror, with an insignificant fraction lost to scatter into higher angles.

The plot also shows another interesting feature, which is that the peak value (at 25mm) is larger for the reference mirror than the Starshine mirrors, by about 10%, indicating a larger reflectance. A more precise measurement was made, and it was discovered that the Starshine mirrors have a reflectance of 89-90% at the test wavelength of 532nm.

For the best results, the Starshine mirrors must be as flat as possible – preferably with less than 5micrometers (10 waves of light) deviation from a perfectly flat surface. The mirrors were tested using an interferometer which combines the light reflected off a perfectly flat mirror with that of the light reflected from the surface of our test mirror to give a contour pattern of the surface variations of the test mirror. The results for these tests are shown in Fig. 3 below. Initially, we tested a "perfect" reference flat that produced straight fringes with little variation, indicating a surface flat to about a tenth of a wave. A series of mirrors were tested with many of them showing a slight astigmatism (indicated by the X-like appearance of the fringes), but overall, the surface errors were between 0.5 and 2 waves, which is well within the desired tolerance limits.

A more serious concern is the effect the mounting scheme has on these mirrors. Since they are to be clamped firmly to a large aluminum shell, it is important that the stress on the mirrors doesn’t deform the mirrors beyond the acceptable 10-wave limit. The effect of mounting stress was examined by making an interferogram of the mirror surface: while free-standing and while mounted to a test plate. The images are shown in Fig. 4 below, indicating that less than 1 wave of distortion is introduced in both cases.

Lastly, a qualitative test was made with a mirror mounted to a plate which was then heated by a high power heat lamp. This was to see the effect on the mirrors surface under extreme temperature changes. The results were encouraging – there seemed to be less than a wave or two of deformation introduced, even with temperature changes well beyond that which will be experienced in orbit.

The results of these tests suggest that the Starshine mirrors are well within the calculated tolerances, and should remain so under the thermal and mounting stresses they will be subjected to throughout the mission.