Speaker
Description
Located in the heart of the Paris-Saclay university and technology park, SOLEIL is preparing a major upgrade of its machine and beamlines. This project known as SOLEIL II aims to provide a 4th generation synchrotron light source by the end of 2030. The new storage ring will have an emittance of 85 pm.rad which will naturally arise challenging questions about the solutions to be chosen for its alignment. Indeed, simulations carried out by the Accelerator Physics group have shown that the storage ring girders and most of the magnets on a girder must be aligned transversally with a tolerance of ± 50 µm (normal distribution with 25 µm standard deviation and truncated at 2σ). For more specific components such as magnets on matching straight section girders, the tolerances are decreased to ± 20 µm (normal distribution with 10 µm standard deviation and truncated at 2σ). Finally, the 354 m circumference of the machine must be maintained to its nominal value at better than ± 2 mm (normal distribution with 1 mm standard deviation and truncated at 2σ).
The SOLEIL II alignment will be mainly carried out using laser trackers whose commonly accepted instrumental uncertainty is 7 µm + 3 µm/m, making the required 25 µm tolerances only achievable under very specific conditions. However, for the 10 µm tolerance needed for the most sensitive components, the limits of the instruments must be pushed back.
All the parameters having a strong influence on the measurement process during a survey must be perfectly controlled and optimized to maintain a low level of uncertainty. This includes the instruments’ metrological performance, the environment in which they are used, how the operators use them, the dimensional and geometrical stability of the machine, and finally the methodology used.
Although an experimental approach is possible for this optimisation, it remains difficult to implement due to the number of scenarios that need to be tested directly on the machine to draw conclusions on the large number of variables that influence the results.
For this reason, the approach envisaged here is to model the measurement process which involves defining its input parameters such as uncertainty components. This model will be fed with experimental data. It will be used to perform simulations and thus identify the best measurement strategy for the future machine. We will optimise a selection of parameters such as the number of laser trackers used, the number of fixed points on the walls and their positions, the number of stations and their position, the measurement range, the number of times these measurements need to be repeated, and the duration of the measurement process, etc...
To evaluate the performance of the different tested scenarios, probabilistic techniques such as Monte Carlo approach will be used. This consists in carrying out random draws in compliance with the conditions of each of the input parameters. The whole of this approach will enable us to identify the uncertainty components having the greatest contributions during a machine survey, to optimise the measurement process and to establish the best geodetic network before installation of the future machine.
The poster will explain how the model is built, what are the main parameters and uncertainty components considered and finally will present the very first results obtained.
