next up previous contents
Next: Symmetry Up: Point-wise Data Previous: Note on Removing

Repeatability and Hysteresis

The raw data from run CW25 for the inclinometer channels (1, 2, 4, 5, 7, & 8) are shown in Figs. 4 & 5. Reassuringly one (CW) dataset looks much like another, particularly after RLB (the removal of assumed linear baseline drifts). Hereafter we are concerned more with their differences and the statistics thereof, and will not reproduce these data graphically unless they contain something of particular note.

After doing RLB (turning OUTCW1.dat into ZCW1.dat, for instance) the datasets were intercompared. There are obvious 'acceleration' effects seen in the first few (say, 10) recorded data of each run, and these were removed before the analysis :

  Data                   Ch   1       2       4       5       7       8

OUTCW1 - CW25     (CW-CW)    0.05   0.29    0.36    0.15    0.23    0.12 " rms
OUTCW1 - OUTCCW1  (CW-CCW)   0.13   0.28    0.79    0.58    0.53    0.57
The differences, CW-CW and CW-CCW, show no spikes, indicating that at this resolution and with this method even the track bumps are repeatably measured. The difference plots (eg OUTCW1-CW25) are essentially noise, and whatever structures lie above the noise are limited in magnitude to 0.3". There is some small (<1") systematic structure in the CW-CCW plots which might be called hysteresis.



Iain Coulson
Thu Mar 13 13:57:23 HST 1997