|
|
| Home | Problem Reports | Operations Schedule | Weather | Data Archive |
|
|
Grant M. Hill January 3 2000 General Comments For the last few months, work on the stacking code has largely concentrated on changes to either make life for the TO's a bit easier or on changes which have the potential to improve stack quality in the future. The TO is now asked to enter the mean truss temperature (MTT) as the code is running. This number is currently not being used for anything but gets stored in the align file header. This is allowing me to archive the align files according to the change in MTT since the previous alignment. The hope is that once statistically large enough numbers of align files are accumulated some functional dependence of tip and tilt correction on MTT change can be searched for and code developed which will allow thermal corrections. The hysteresis database management code presently looks for a dependence of hysteresis on MTT and can generate coefficients to describe such a dependence should it be found. The stacking code is now able to take these coefficients and use temperature dependant hysteresis corrections instead of straight averages. Presently this capability is turned off though because we do not have a statistically large enough number of corrections.
Life for the TO's has been made a bit easier by significantly
reducing the number of mouse and keyboard clicks when the manual
version of the stacking code is run. Audible warnings are now
issued when incorrect numbers of mirrors are found. When things
don't go the way they should, exit from the code can be accomplished
much more gracefully now. During the time covered by this report (Oct, Nov, Dec 99), stacking speed remained roughly constant. Measured from arrival at CCAS to departure from CCAS the median time is about 10 minutes.
Until now, I have been presenting plots of stacking speed in seconds per
mirror. This was intended to emphasis progress in speed that
may not have been obvious as mirror number increased. Since we have
now been using all 91 mirrors for about a year and have now entered
into early operations, it is more meaningful to show actual time
taken to stack and time away from science. The upper panel of
Figure 1 now shows time from arrival at CCAS to departure from CCAS
(circles) and time away from the sky (crosses). The lower panel of Figure 1 shows the
change in EE(50) and EE(80) over the last two years binned into
approximately one month bins.
The slight increase EE(50) and EE(80) during the month of December
is probably due to degraded hysteresis corrections and actuator
failures due to cold weather.
Figure 2 presents a histogram of EE(50) and EE(80) results
during the months of October, November and December.
Top of Page Return to Technical Reports List |
Created: see
report date above Send comments to: webmaster@het.as.utexas.edu |
