Number of telescope cycles using unbalanced moment
02mar11
Draft
Intro:
The tension in members of the telescope changes as
we move the dome, ch, azimuth, and tiedown tensions. The lifetime of a
beam is counted in the number of times the change in tension goes
through a cycle. If the change in tension during the cycle remains less
then 16Kips/sqIn then the cycle does not affect the lifetime of
the
member.
Amman and Whitney (21may10) provided a table of stress ranges for members
over 100% capacity (.pdf). Using their simulator, they moved the
telescope through a range of motions and looked to see how the stress
changed in a particular truss (I'm not sure what they used for tiedown
tensions in the calculation).
The following sections contain:
The dataset: (top)
The telescope position is recorded once a second to
disk. Some properties of this dataset are:
 Data includes 2004.5 through 2010
 There are a few days of missing datafiles (disc problems etc.).
this is less than 1 % for any given year.
 There are times when the data samples are invalid (time,
position..). This can occur during an encoder failure or when
part of the system is powered down to be worked on. This data is not
"missing" since the telescope can not move when this maintenance occurs
(so no cycles can occur).
 2007 was the platform painting project. The telescope motion was
stationary for a large part of this year.
Examples of telescope
motions (top)
The plots show the az, za, and unbalanced moments
for various observing modes used at the telescope:
 Aeronomy azimuth swings (.ps) (.pdf):
 Page 1: 14jun05: dome=15,ch=15, unbalanced moment +/
9500KipFeet.
 Page 2: 08sep05: dome=15, ch=0, unbalanced moment:+/
13500 KipFeet
 both of these used .4 degrees/second in azimuth.
 About 95 (360 degree) cycles per day.
 Number 2 is the typical run mode.
 Driving dome up and down in za at
az = 0 degrees (galfacts) (.ps) (.pdf):
07dec08
 az=180 degrees, ch=8.8, dome=2 to 19.6 degrees.
 Unbalanced moment 19500 to 11000 KipFeet.
 this takes 20 minutes to go up,down.. 3 cycles per hour and
runs for 4 to 6 hours a day.
 3 days of typical tracking (.ps)
(.pdf) 2325apr05
 top: azimuth position
 middle: dome position
 bottom: unbalanced moment at td 12.
 The large swings in moment come from swinging the azimuth arm
through 360 degrees with dome at 19.6 degrees.
 the typical use of the telescope looks like these days rather
than the repetitive motions of the first two plot.
Number of
cycles using change in moment: (top)
The number of cycles on the telescope was computed
using the change in the unbalanced moment (at a corner). The Amplitude
of the change for a cycles was computed for changes greater than:
[5000,10000,15000,20000] kip feet.
The computation constisted of:
 Compute the unbalanced moment on the azimuth arm as a function of
time (from the dome motion).
 Project this unbalanced moment on the 3 corners of the triangle
(0,120,240) degrees:
 UnbalMomentCorner=UnbalMomentArm * cos(azimuthCorner
azimuthOfArm)
 For each corner process all the data (once a second sampled)
counting the number of cycles performed.
 A cycle consists of:
 The (Max  Min moment seen surpasses the specified
amplitudeDifference (ampDif).
 It must then go in the opposite direction (once the first
ampDif is found)
 It should then return by +/ ampDif/2
 The order of the difference (increasing the decreasing or
decreasing then increasing) is handled by the program.
The plots show the change in unbalanced
moment on the
telescope 2004.5 through 2010 (.ps) (.pdf)
using
the above definition:
 Page 1. the number of cycles by year
for each corner and the 4 Amplitudes
 The cycle amplitudes used were:

Frame 
Amplitude
(kipFeet)

1

>
5000

2

>
10000

3

>
15000

4

>
20000

 For each frame Black is corner 12, red is corner 4 and green is
corner 8.
 The dashed blue line shows the yearly median counts (excluding
2004,2007, and 2010)
 Year notes:
 2007 is the platform painting year which had limited motion.
 2010 had the za limited because of the broken beam repair.
 Page 2: Integrated cycles 2004.5 thru
2010 for the 4 cycle amplitudes.
 The total cycles for 2004.5 thru 2010 is plotted versus
cycle amplitude.
 the 3 colors are corner 12,4,8.
 Page 36: Histogram of cycle
depth for each year
 Each line in a frame is a histogram of the cycle depth for a
given year.
 For each cycle counted, the program ,kept track of the actual
cycle amplitude.
 The histogram binwidth is 800 kipfeet.
 Each frame has 7 histograms (one for each year).
 The 3 frames on each page are corners 12,4, and 8.
 Each page is shows a different Cycle amplitude :
5000,1000,15000,20000 KipFeet change.
 Some bins with spikes probably correspond to repetitive motions
of the telescope.
Summary:
Number of telescope cycles using Amplitude of unbalanced moment
change:
Number of telescope cycles
integrated 2004.5 thru 2010
Amplitude
KipFeet

corner12
totCyles

corner
4
totCycles

corner
8
TotCycles

> 5000

48665

53061

52568

> 10000

29838

28155

27620

> 15000

15469

15536

14985

> 2000

8320

8773

8384

 The above table shows the accumulated cycles for the 6.5 years.
Median Cycles/year
Amplitude
KipFeet

Median
Yearly
cycles

>
5,000

9378

>
10,000

5418

>
15,000

2980

>20,000

1704

 The median cycles/year has excluded the 2004,2007, and 2010 data
Unbalanced moments for typical motions:
 Some of the motions used on the telescope:
 World day az swings:
 dome=15,ch=15 = 9500 Kipfeet.
 dome=15, ch=0 = 13500 kipFeet.
 Galfacts za nodding
 az=0, dome 2 to 19.6 deg
 max unbalanced moment: 18500 KipFeet
 min unbalance moment:  9000 KipFeet
 swinging from source set to
source rising
 will have a moment of
+/ 19500 kipfeet *cos(azRise,set  corneraz)
 it can take 1.5 to 2.5 hours to track a source rise to set.
 The tiedown tensions were not used at all in this computation.
processing:x101/110204/agccyclesdif.pro,
agccycleslev.pro
home_~phil