Thermal modelling, conception and design of a cooling system for the lhc injection kicker magnets towards high luminosity lhc operation

Resumen

The CERN Large Hadron Collider is equipped with two fast single-turn injection kicker systems that deflect the incoming particle beam onto the accelerator’s orbit. The high intensity LHC beam, circulating for many hours, can cause considerable heating of the injection kicker magnets. If the Curie temperature is reached at the magnet yokes, made of ferromagnetic material, the next beam injection in the LHC has to be postponed until the yokes are cooled to avoid mis-injection of the beam, affecting the whole LHC operational schedule. Thanks to the studies performed in the context of this thesis, it was possible to gain knowledge about the amount of power deposited and its distribution along the different components of the magnets due to the interaction with the circulating beam. This was key to propose a cooling method. Finite element models have been developed with ANSYS for thermal analysis, and validated by benchmarking with measured temperatures during LHC operation. These models are used to predict the temperatures reached for various operation scenarios. According to these predictions, heating issues are expected for High Luminosity LHC (HL-LHC) operation with high intensity beams unless appropriate measures are taken. Several alternatives for cooling the magnet yokes have been investigated, always taking into consideration the limitations imposed by the environment of these magnets: They are operated in ultra-high vacuum conditions and pulsed at high voltage. Due to this, at the initial stages of the studies, a passive cooling method was pursued due to the absence of operational risks associated to its installation in the magnets. In the context of the passive cooling techniques, the investigation was directed towards improving the radiative cooling inside the vacuum tanks that contain the magnets, by increasing the thermal emissivity of the internal surface. The results from simulations confirmed that this approach would reduce the temperature of all the components for HL-LHC operation, but the first three magnet yokes at the entrance of the magnet would still be above their Curie temperature. However, given the intrinsic advantages of the passive cooling, it was kept as the core solution while other complementary measures were investigated. Hence, a test campaign was launched with different companies and institutions to find a suitable treatment, and two coatings based on thermal spraying processes were found to be suitable thanks to their adaptability and the good results obtained. However, by the time the vacuum compatibility tests, compulsory for installing any new material and coating inside the LHC vacuum, were finished, the studies of the complementary measures were well advanced, and some approaches were showing promising results, even without the need of increasing the emissivity. Although a liquid cooling system was initially discarded due to the above-mentioned limitations, simulations showed that it was the most efficient option. The investigation was then oriented towards the implementation of a reliable cooling system operating safely in the LHC vacuum environment. Beam coupling impedance studies were performed to study the possibility of relocating a significant portion of the heat deposition from the yokes to ferrite rings placed at the entrance of the magnet, and thus, more easily accessible, which are not critical components for the functionality of the magnets during beam injection. Based on these studies together with extensive thermal analyses, it was shown that a reduction in the temperatures of all the components could be achieved for HL-LHC operation by redistributing the power deposition plus installing a cooling system in the rings. The conception and design of the cooling system in the ferrite rings was particularly challenging due to the difficulties imposed by installing a liquid cooling system in an ultra-high vacuum environment. Besides, ferrites are extremely difficult materials for heat extraction due to the intrinsic poor mechanical properties and limited thermal conduction. An optimization campaign of the design was carried out by means of coupled thermo-structural simulations. Besides, given the criticality of the application, simulation tools were implemented to import accurately the power distribution obtained with CST, the software used for electromagnetic simulations, to ANSYS, the one used for the thermo-structural studies. This was key to properly assess the thermal stresses generated due to temperature gradients between the cooled surface and the hot areas in the ferrite rings. Different alternatives were also investigated to ensure a good thermal contact between the ferrite ring and a cold plate of copper used as heat sink. In the end, a brazing technique was applied. A test campaign was launched to prove the feasibility of this technique given the high mismatch of the coefficient of thermal expansion of both materials. After some iterations, the technique was optimized and a sound joint was achieved. A full prototype of the cooling system has been built and assembled in one kicker magnet, which will be installed in the next period of operation of the LHC to prove its reliability during operation with high intensity beams. If it is proven to be successful, all the kicker magnets will be upgraded for HL-LHC, including the cooling system proposed in this thesis. lution.