Prior to June 1995, our track models had a resolution of 1
. Tracking
across the worst joints was compromised, as shown in Fig.11, since the
track profile varies on scales more typically 0.2
by as much as
5
.
Inclinometry and track models since then have been obtained with this higher
resolution within 1
of azimuth of each wheel-joint interaction.
Tracking experiments since then, over the worst joints, show no discernible
pointing errors on the strip chart record. We estimate that the residual errors
might be 0.5
across each joint. With 1
-wide joints
covering
15% of azimuthal space, the contribution to the error budget might be reckoned
as 0.1
rms.