Processing magnet geometry measurements for better control of LHC aperture

The large hadron collider superconducting axes of the magnet are measured from both ends. These two redundant measurements are combined to get a reliable measurement result. When the two measurements are put together, we observe a "saw tooth" effect due to the fact that the two measurements are, in general, not identical. This is expected from the accuracy of the two measurements. However the effect observed, in the vertical plane, is considerably larger than expected. Effects of temperature gradients in the cold bore tube during measurements have been observed and we show that this effect is the most probable explanation for the observations of the large differences in the measurements between the two sides. This work proposes an algorithmic approach to filter this effect to improve measurement results. Magnets are positioned with an accuracy of 0.1 mm, and the error in positioning coming from measurement errors due to the temperature effects can be up to 0.3 mm. Our analysis shows that by applying this correction we can assure the best positioning of the magnets in the tunnel in the vertical plane. Analysis is done for the 15 m long main dipoles, for which the effect is most visible.