BBA Results

In order to have a well defined tune variation ($\delta\nu$ = $\pm$0.025) during the beam-based alignment measurement, the previously measured average beta functions (see section 1.1) are used to determine the allowed change of quadrupole strength. A hysteresis correction restores the orbital tunes after each quadrupole variation cycle, in order to minimize the residual distortions of the linear optics.

Figure 3: BBA data for BPM ARIDI-BPM-07ME showing a vertical BPM offset of 47.5 $\mu$m with respect to the adjacent quadrupole ARIMA-QMD-07 at a beta function of 18 m

Fig. 3 shows the result of a single vertical BPM offset measurement for ARIDI-BPM-07ME. After taking a reference orbit, the adjacent quadrupole ARIMA-QMD-07 is changed by $\Delta kl$ = 0.017 m$^{-1}$ followed by a variation of a local orbit bump of $\pm$0.4 mm. The square of the standard deviation of the difference orbit, excluding ARIDI-BPM-07ME, versus the BPM reading is fitted by a parabola. The difference between the minimum of the fit and the zero reading of the BPM determines the BPM offset. In this case the measurement reveals an offset of 47.5 $\mu$m within an error of $\pm$0.5 $\mu$m. Fig. 4 summarizes the result for 66 vertical BPM offsets with measurement error variations between $<$ 1 $\mu$m and 50 $\mu$m. The offset distribution is fitted by a gaussian shifted by $-$0.11 mm with a standard deviation of 0.24 mm. Three BPMs show offsets larger than 0.5 mm.

Figure 4: Vertical BPM offsets with respect to adjacent quadrupoles in the SLS storage ring. The offset distribution is fitted by a gaussian shifted by $-$0.11 mm with a standard deviation of 0.24 mm. The measurement error varies between $<$ 1 $\mu$m and 50 $\mu$m.

The offsets have been fed into an SVD based global orbit correction code in order to determine the corresponding corrector pattern. These corrector predictions (`` kick change'') can be compared to the actual corrector settings (`` present kick'') for a flat orbit. Fig. 5 depicts both patterns and their difference (`` sum kick''). It can be seen that they are ``anticorrelated'' resulting in a 20 % reduction of the rms corrector kick from 0.15 mrad to 0.12 mrad. Furthermore the mean corrector kick of $-$0.014 mrad is removed. The application of the vertical BBA data in the storage ring confirmed the expected rms kick reduction.

Figure: Comparison of actual corrector settings (`` present kick'') for a flat orbit and predictions for the correction of the vertical BPM offsets (`` kick change'') predicting an rms kick reduction by 20 %. The squares (`` sum kick'') denote the differences