In reality, however, the track is irregular in profile. As the antenna
moves over it the 4 supporting wheels will each move up or down, particularly
with respect to the other wheel on the same side of the antenna. The two
wheels on each side are part of an A-frame, whose apex is one end of the
elevation axis. Motion of the wheels thus causes that end of the elevation
axis to rock up or down, forward or backward, in the plane of the
A-frame.
This causes the elevation axis to be moved in the (aero-)nautical directions
of roll (the right end of the elevation axis rises w.r.t. the left, or
vice versa) and yaw (the right end moves forwards w.r.t. the left, or
vice versa).
The top of the A-frame also rotates on an axis
perpendicular to the plane of the A-frame, which would cause the elevation
axis to pitch if it was fixed to the top of the A-frame. However,
the elevation axis is free to rotate within the bearings atop the A-frames,
and its orientation is determined using the encoder fixed to the right-hand
A-frame. During normal observing the SERVO system is ON, allowing for
orientation of the elevation axis w.r.t. the right-hand A-frame. Pitch
affects the elevation pointing only.
The track comprises 14 segments, of two sizes (see the notes to Table 1),
which have their own peculiar profiles, abut imperfectly, and sit atop a layer
of grout whose longevity is questionable. The track profile as a function of
azimuth must be described by a look-up table at sufficently high angular
resolution. The profile is likely to change with time,
either gradually, as the grout crumbles beneath it, or abruptly, following
a change in loading, or following an earthquake, for instance. The profile
must necessarily be measured in the advent
of any such event, and, in any case, on a regular basis.
The track segments were surveyed in 1991 by a crew from the SERC Council of
Works unit in Swindon and showed deviations from planes
of about + 0.015 inches, with abutment errors typically
0.005 - 0.010 inches. Inclinometry measurements made of the (8-metre long)
beams joining
the front and back legs of the antenna show corresponding tilts of about
+ 10" as the antenna moves in azimuth, although there is
probably a dynamic interaction between the antenna wheels and the track which
prevents a strict correlation between the (static) track profile
survey measurements and the inclinometry measurements made at the same
azimuths. These effects propagate
to errors in the pointing of the telescope of roughly the same magnitude,
10".
The inclinometry project aims to correct the pointing of the JCMT for the
effects of the track profile, and, as part of the Pointing Project,
to keep pointing errors in azimuth and elevation to 1" r.m.s.
( Pointing Project Meeting No.1, 940201)