First (Light?) Test...
Usage of the Foucault tester started
in February. While the mirror was not really ready for testing, it
was satisfying to test the mirror and see a dim pattern in the expected
form. The room is darkened since the image is quite dim. (Remember,
it has not been aluminized yet!) We had to make sure to warn
both Jim and John before the flash went off, since eyes need to be adapted
to low light levels to see the reflected light.
Jim is seen trying to find the reflection of light from the slit on the
Foucault tester.
John is seen adjusting the mirror stand up and down by adjusting the screw
in the front of the holder. By moving the entire cart side to side
and the adjustment screw up and down, the reflected image from the slit
should be focused back to Jim and the knife edge.
Once the knife edge is properly placed, this is the image that was seen
when looking through the tester. A spherical surface will appear
flat. There is still some surface roughness left since the mirror
is not yet fully polished. The raised ring is due to some problems
with the curvature of the prepolishing tools. Fortunately, this is
(very roughly) the general appearance that a paraboloid shows in the test,
although it shouldn't be as extreme as in this appearance.
Notice that you see the inside of one ridge on the mirror and the outside
of the other. The bright lines down the centerline of the image are
actually reflections from the back of the mirror, we believe. If
you look carefully at the image, you can barely see a pattern of spiral
lines. These are from the periodic strokes of the grinding machine.
Since this is early, this is of little concern. Later polishing should
remove this error.
A ray tracing of what is happening will help to explain the image above.
The ridge is greatly exagerated to show how a parabolic surface varies
from a spherical surface of a particular radius of curvature (in fact,
the surface is certainly still all convex, but is raised at a certain distance
from the center, relative to a sphere) . In this example, the light
rays are shown as if they originate at the knife edge, which is essentially
true for some Foucault testers (but not quite for ours, where the light
comes from a point a bit to the side, but that makes little difference).
It may be easier to follow this if you think of the rays as coming from
a point rather than a slit (the only reason a slit is used is because the
light source would have to be very bright to get enough light through a
pinhole, although a laser focused through a converging lens would solve
that problem). The black lines are light rays reflecting off a portion
of the mirror that is spherical, so they focus back to the knife edge and
are essentially half-blocked. Thus those parts of the mirror appear
gray. The blue lines are rays that will hit the knife edge and therefore
not be seen. Those parts of the mirror will appear black. Notice
that for one side, the inside of the ridge is cut off by the knife edge.
For the other side, the outside is cut off. The red lines are rays
that will pass the knife edge and make it to the eye for detection.
Those parts of the mirror will appear bright. The result appears
almost the same as an extremely exaggerated relief view, as if the light
was coming to the mirror from a very low angle (the only difference is
that the "ridges" don't cast shadows, but one side of each ridge is still
dark). The "ridge" in the photo above is only on the order of 1/100,000
of an inch high.
Second test...