In September -99, it was discovered that there was significant motion of the images in the LiF 2, SiC 1, and SiC 2 channels relative to the LiF 1 channel, caused by the varying thermal environment of the orbit
As guiding is done with the LiF 1 channel (maintaining the guide star image locations with respect to the slits), motion on this channel is largely eliminated by the ACS
Primarily, motion of the LiF 2 is in the Y direction (along the slits), and SiC 1 and SiC 2 motion is in the X direction (across the slits)
Possible Causes
An engineering study has been done on the image motion, and has identified two probable causes for this motion
“Kick plates” which allow access to the mirror cavity (used for maintenance activities during integration and test) are overly constrained to the mirror bench. Changes in the plate temperatures applies a moment to the mirror bench, causing the mirrors to rotate with respect to each other
Unintended coupling may be occurring between the aluminum baffle and the mirror bench. The baffle assemblies are long (from their mounting points to the mirror bench) and are thermally uncontrolled (the tops of the baffles are known to vary 30 degrees C or more with variation of pointing)
Typical Target to Target Motion
When slewing from target to target, the following motions are noted
LiF 2 Y motion of approximately 0.4 arcseconds per degree beta change (angle to anti-sun vector)
SiC 1 and SiC 2 X motion of approximately 0.5 arcseconds per degree beta change
SiC 1 and SiC 2 Y motion of approximately 0.2 arcseconds per degree beta change
SiC 1 and SiC 2 X motion of approximately 14 arcseconds when going from CVZ to non-CVZ target or vice versa
Typical Orbital Motion
Over an orbit, the following motions are known
LiF 1 X motion of ~4 arcseconds full amplitude, break points at dawn and dusk for CVZ targets (fairly flat through day or night)
SiC 1 and SiC 2 X motion of ~3 arcseconds full amplitude over the orbit for CVZ targets (fairly flat through the day, spike at night), non-CVZ targets appear to have significantly larger orbital motions, perhaps as much as 6 arcseconds full amplitude
Additional characterization currently in progress
Coalignment Maintenance
“M112s” run on an as-needed basis
scans in X and Y to measure each channel's location
mirror and/or FPA motions then made to coalign
Predictions of image locations throughout an MPS are made
primarily FPA motions used to correct for anticipated motions (occasionally mirror motions as well)
Peak-ups used to maintain alignment in certain cases
Observations checked both "on the fly" and afterward to close the loop
Possible Effects on Observing Programs
Targets observed immediately after large beta change, or CVZ/non-CVZ change risk the loss of one or more channels in the LWRS
A typical scenario is the loss of LiF 2 and/or both SiC channels
Channels may reappear in later exposures (assuming adjustments made)
Targets may have motion within their slits over the course of an exposure
Correction would be similar to what is done with spectral motion, except that airglow cannot be used as a reference
MDRS slits
LiF 1 and LiF 2 for all targets with minimal risk
All channels for CVZ targets with moderate risk
SiC 1 and SiC 2 for non-CVZ targets not recommended at this time, expect to have more information by end of February