Azimuth bending errors dec07

• Page1: The first 2 plots show the az (dome side) and za (dome) for the two events (the dashed green lines show the events). The bottom plot is the difference in the two encoders (enc1-enc2) vs hour (the units are encoder counts).
• Page2: The first event at 5:42.
  
• Top: A blowup of the azimuth position (dome side in degrees) at the failure.
• middle: encoder difference at the failure.
• Bottom: The values of the two encoders at the failure.  Black in encoder1 (dome side) and red is encoder 2 (ch side).
  
• Both encoders continued to move during the failure so they remained in contact with the rack gear.
• enc2 did not move as fast as the enc1. This error looks like a true bending error.
• Page3: 2nd event at 7:42 am. When the encoder was inspected, the encoder gear was not engaged in the encoder rack gear. Pushing on the spring loaded encoder mount engaged the encoder in the rack gear.
  
• Top az position vs hr. The azimuth was increasing clockwise.
• middle: encoder difference (enc1-enc2) vs hour.
• bottom: enc1 and enc2 positions.
• encoder 2 (ch side) stopped moving at the failure. It was no longer engaged in the rack gear. The encoder rack gear is not perfectly round (.ps) (.pdf) . Looking at the az encoder difference plots, the bending failure at az=327 degrees is a spot where the enc1-enc2 is going negative. The enc2 side of the rack gear is bowing out (enc2 is moving along a longer arc than enc1). Something prevented the encoder mount from pushing out far enough to remain in contact with the rack gear.

• The azimuth was moving counter clockwise at slew rate with the dome at 10 degrees.
• The top plot shows the azimuth position. The bottom plot shows the encoder difference.
• The red enc2 (red, ch side) stopped moving. When the encoder was inspected it had disengaged from the rack gear by about 1/2 inches.
• This is not a spot where the rack gear curvature changes rapidly.

Conclusions:

• The az enc2 (on the ch side) has come out of the rack gear in two spots (az=327 and az=273). The rack gear bows out at 327 but not so much at 273.
• After the  az=327 failure, the encoder cover was found to stop free motion of the encoder. After fixing this we still had the az=273 failure the next day. We need to disassemble the encoder mount and clean it.
• The error at az=53 was a true bending. When we drove the azimuth back and forth over this region, there was a vibration in the platform. It was coming from the sides of the azimuth wheels rubbing against the azimuth track. This can cause a bending error (especially if we are reversing directions at this location). The azimuth rails no longer have any grease on the top or sides of the rails.We  need to regrease them to prevent these errors.

home_~phil