Recently, Rudy Schild suggested that we perform a Foucault test on the primary mirror. Andy Szentgyorgyi provided a platform, assorted mounting parts, a beam-splitter pellicle and dome-flat lamps. Wayne Peters provided a power supply. Emilio and Rudy mounted the hardware onto the platform, and mounted it on a sturdy tripod to hold the apparatus pointed to the mirror.
The configuration was:
mirror ----- ^ ^ | | : : : : V | pellicle \<----* lamp ~4.6m | | | V eye -----
The eye is placed against a knife-edge (a razor blade) with its edge perpendicular to the beam. The knife-edge must be placed at the center of curvature of the primary (about 4.6m). The lamp and your eye must be at the same distance from the primary.
Visit this link for a nice description of the test if you are curious.
Bob Hutchins helped us with the mechanical setup, which we first tried out on 03/31/04. The 4.6m means that the tripod had to stand on the opened lower dome slit. We managed to aim the lower-lip of the slit over the roof so we could stand on it, after propping it up with wooden blocks and a jack. We managed to avoid putting pressure on the more delicate skin of the lip by standing on its ribs. At dusk, we located the center of curvature of the primary after pointing the telescope at the apparatus. The arrangement was fairly precarious, but we were able to see interesting structure (see below) in the views of the illuminated primary. We decided to repeat the experiment with an improved wooden base with 8-ft lengths of 2x4s and a plywood platform. On 04/08/04, we repeated the experiment. We saw the structures and recorded them again, but we found our arrangement needed to be sturdier, because our own weight would flex the platform sufficiently so we could not set up once and have different observers. We decided to repeat the experiment once more, this time using 4x4s and having equipped ourselves with a digital camera. On 04/20/04, we conducted the final iteration. The results are shown below.
The views above were obtained with a hand-held camera, so they are not ideal, but they do show the structure we saw from the beginning. Different views were obtained by placing the camera lens against the knife-edge and by moving the latter left-right and backward-forward, using micrometer adjustments.
The obscuration from the pod holding the secondary mirror is unfortunately very large. The structure that is clearly visible is reminiscent of the honeycomb shape of the back of the mirror. In addition, some of the shadows appear to have corrugation-like patterns with random orientations, perhaps reminiscent of a lap-tool pattern. A final test that might be conducted with our setup is to obtain similar views when the mirror is in Tucson for re-aluminization in Aug 2004.
The conclusion from these tests is that there may be unwanted structure on the mirror surface. However, it remains unclear whether such structure degrades the seeing to the degree that we measure. The wavefront sensor that we hope will be built this year should provide the final answers.