a b s t r a c tThe future Compact LInear particle Collider (CLIC) under study at CERN will require to stabilize heavy electromagnets, and also to provide them some positioning capabilities. Firstly, this paper presents the concept adopted to address both requirements. Secondly, the control strategy adopted for the stabilization is studied numerically, showing that the quadrupole can be stabilized in both lateral and vertical direction. Finally, the strategy is validated experimentally on a single degree of freedom scaled test bench.

a b s t r a c tThis paper presents an experimental validation of a control strategy capable of both stabilizing and positioning the heavy electromagnets of future particle colliders. The originality of the approach is to use the same active mounts to perform both tasks, with a nanometer precision. In a previous paper, the concept has been studied numerically, and validated on a scaled single degree of freedom (d.o.f.) test bench. In this paper, it is extended to a two d.o.f. test bench, constituted of a heavy mass mounted on two active legs. Firstly, the model is described and the performances are discussed numerically. Secondly, experimental results are presented, and found to correlate well with the model, and comply with the requirements. Finally, the experimental results are combined with a simplified model of the beam-based feedback to evaluate the jitter of the beam. It is found that, at the scale of a single quadrupole, the mechanical stabilization of the quadrupoles reduces the vertical beam jitter by a factor 10.

To reach a sufficient luminosity, the transverse beam sizes and emittances in future linear particle accelerators should be reduced to the nanometer level. Mechanical stabilisation of the quadrupole magnets is of the utmost importance for this. The piezo actuators used for this purpose can also be used to make fast incremental orientation adjustments with a nanometer resolution. The main requirements for the CLIC stabilisation electronics is a robust, low noise, low delay, high accuracy and resolution, low band and radiation resistant feedback control loop. Due to the high number of controllers (about 4000) a cost optimization should also be made. Different architectures are evaluated for a magnet stabilisation prototype, including the sensors type and configuration, partition between software and hardware for control algorithms, and optimization of the ADC/DAC converters. The controllers will be distributed along the 50 km long accelerator and a communication bus should allow external control. Furthermore, one might allow for an adaptive method to increase the S/N ratio of vibration measurements by combining seismometer measurements of adjacent magnets. Finally a list of open topics, the current limitations and the plans to overcome them will be presented.

No abstract

No abstract