The earthquakes.

    Puerto rico experienced a number of strong earthquakes starting on 6jan20 through 11jan20 (see a list of quake, strength, cable/tilt sensor oscillation freq). After the first quake on 6jan20 the telescope was parked with the motors off and the brakes on (except for one or 2 instances).  The az encoder (read once a second) showed that the azimuth arm moved during each of the stronger quakes. This is to be expected since:

• The azimuth flanged wheels sit on the azimuth rail.
• the brakes will keep the wheels from turning, but the wheels can still slide on the azimuth rails if the force overcomes the static friction between the wheels and the rail.
  
• The azimuth arm rotates about a central bearing. We need to be careful that there is not too much bending in the azimuth arm  because of some differential motion in the carriage house and gregorian dome sides of the azimuth arm.
• the carriage house and the dome do not have this problem since they are pulled up and down the azimuth arm via a geared track.

Some occurrences that may have moved the telescope (times are ast epicenter times):

Plots were made showing the motion of the telescope 6 jan to 28 jan20. (.pdf)

• Page 1: azimuth positions, az encoder differences 6jan to 28jan20
• top: azimuth position
• The times when the azimuth moved is flagged with an adjacent label with the strength of the quake.
• The largest jump was the 5.9 quake of 11jan20, where the azimuth moved more than .1 degrees.
  
• After a power outage, the azimuth value can drop off the page (system needs a reset).
• bottom azimuth encoder difference (azgr - az ch).
• The largest jump was the 5.9 quake of 11jan20. It moved about .01 degrees.
• Page 2: greg dome and carriage house za position.
• Top: dome za.
• The 6.4 quake showed a .03 deg change in the encoder.
• This turned out to be the operator entering the pntpwrdip command after the power came back. This turns on the motors and caused the dome to move by .03 degrees.
• The only other dome motion was on 22jan20 around 18:00. The observer started cima and moved the telescope .03 degree (to the position it had on 06jan20). After about 20 minutes they moved it back to the 8.03 deg position.
• bottom: carriage house za.
• there was no motion of  the ch.
• Page 3-4. these are blowups of the az position and az encoder difference for the stronger quakes.
• page 3 frame 1,2  - 07jan20 6.4 quake.
• the power dropped out for about 6 minutes (the generators came on and it took the operator awhile to reset things.
• Looking at the motor status i saw that the dome motors were turned on for a few second but the azimuth motors were never started.
• page 3 frame 3,4 07jan20 5.6 quake
• the azimuth moved by about .06 degrees.
• page 4 frame 1-2 5.2 quake 10jan20
• no motion of the azimuth position.
• page 4 frames 3-4 5.9 quake 11jan20
• the azimuth moved by .14 degrees
• the encoder difference changed by about .01 degrees .. so the azimuth bending changed a little
• Page 5 blowup of dome za
• there was a .03 degree move on 07jan20 caused by the operator
• there was a -.03 and then .03 move on 22jan22 caused by the observer (lasting about 20 minutes)
• Other than these 2 moves, the dome remained in the same position to within .0006 degrees
• Using a radius of 420.75 feet (128.245m) from the center of curvature to the rolling surface,  a .0006 degree motion is 1.3 mm

SUMMARY

• With the motors off and the brakes on, the azimuth slid by up to .14 degrees during some of the earthquakes
• using a 60 foot radius .14 degrees is 1.76 inches or 4.5 cm.
• there was a slight increasing in the bending during one of the motions (.01 degrees).
• The azimuth position is measured with an encoder that rides on an azimuth wrack gear.
  
• As long as the encoder remains engaged with the rack gear, there is no loss of position information.
• On 30jan20 osvaldo and willie checked the azimuth position using the scribe mark pointer (at az=270). They found that the encoder value was correct (there was no jump of the encoder).
  
• the gregorian dome and ch did not slide during the quakes since they run on a geared track.
  

processing: x101/200130/telposquake.pro