|
|||
| Joint Astronomy Centre |
|
|
|
| Show document only |
|
||||||||||||||||||||||
|
____________________
|
||||||||||||||||||||||
|
JCMT central bearing Dial Gauge measurementsVertical deflections as measured between soleplate (static) and rotating part of central bearing. Following up on an idea generated on 6/7th May I performed some measurements on the JCMT central bearing using a dial gauge indicator to look for physical displacements betewwn the lower (fixed) and upper (rotating) parts of the housing. The idea was that if we indeed have indents in the race you would be able to see the upper housing drop in to them. Detection of central bearing flaws was quite clear at all known locations, yielding deflections of about 20-40 microns with a width of around 3 degrees. This was consistent with the existing empirical model given that my azimuth errors were larger and the displacement only becomes visible at effects of > 5-10 microns. The set up was with a dial guage capable of resoving about 1/4 thou (6 microns) and azimuth readings were coordinated by walkie-talkie with the observer (GMS) during Friday's Mars observing. Extremely interestingly, the effect was not in the sense I expected (ie falling into holes) but was visible as a climb of the top half of the casing. To check I hadn't made some 14,000ft error I phoned RJSG to talk through my methodology: we agreed I had done this properly. He also suggested I move 180deg around the casing as a check : I duly moved to the south side and found that the casing was dropping on that side, by about the same magnitude. We had to leave shortly after that so I was not able to obtain further points at other orientations (so may have missed the maxima). So, pending further investigations we have an effect which also permits a tip of the central bearing: what does this mean? Several options are available. I would tend to dismiss a single 'big lump' or trough as this would not very well explain equal sized tilts of the housing in opposite senses 180 deg apart in azimuth. My first thought was to postulate a set of five or six matched brinelled 'dimples' on one side of the bearing caused by an instantaneous event (earthquake, electric arc or servo chatter). These would line up in the top and bottom races at one unique azimuth, but when the telescope rotates the top set would move around to the other side: hence permitting some tip. The effect on pointing is entirely due to torque, so probably does not care too much which particular top or bottom flaws are lined up at any given time. To first order I would imagine the tip of the bearing will have negligible effect in elevation as the motion does not get transmitted via the A-Frame members. This may explain the non-linear baseline noted by IMC in the latest data though.
calibration for North data :
Note : North DTI location was 12:45 deg eastwards from true North
calibration for South data :
Note : South DTI location was 167:57 deg eastwards from true North
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
