CERN Accelerating science

In this thesis, research and development work for the air cooling system of the Inner Barrel of the Inner Tracking System 3 (ITS3 IB) is carried out. The ITS3 IB will be a future particle detector of the ALICE experiment at the Large Hadron Collider (LHC). Efficient cooling is crucial for maximizing the particle measurement quality. The cooling system introduces open-cell carbon foams as heat exchangers that combine conduction and convection for the first time in a high-level engineering application. The ITS3 IB will have an unprecedentedly low material budget, which is a requirement for particle detectors that aim at precise tracking at low particle momenta. In the first chapter, the characterization of open-cell foams is performed. A multiscale model is developed that is applicable to a wide range of parameters. The microscopic geometry is based on a periodic model of the foam unit cell. The geometrical scales given by the model are validated with microscope images and computed tomography scans. The outputs of the microscopic model are the pressure loss, the thermal conductivity, and the Nusselt number. These values are used as inputs of the macroscopic model that determines the thermal performance at large scales. The multiscale model is concluded to provide accurate results with respect to the experimental data. The results obtained in the previous study are used to select appropriate foams for the ITS3 IB and to design the cooling system in the second chapter. A prototype of the ITS3 IB is built and installed in a wind tunnel to determine the thermal performance. The analysis also uses numerical simulations. The numerical results show excellent agreement with the experiments, with mean deviations of less than 0.5 K in the temperature distribution attributed to uncertainties in the sensor installation. The results confirm the system adherence to the temperature requirements. Compared to the baseline design, it is shown that enhancements based on simulations can reduce up to half the maximum temperature variation of the detector layers. The airflow applied to the low-mass structure of the ITS3 IB generates vibrations. This requires an aeroelastic analysis, which is done in the third chapter. A novel experimental approach is proposed using confocal sensors to measure the structural displacements of the prototype used for thermal analysis. A mathematical model for the fluid-structure interaction is developed. The results obtained in the experimental setup show that the mathematical model predicts the primary peaks of the spectrum of the structural displacements induced by the air flow. The validated model is used to analyze the real anticipated configuration of the future ITS3 IB, as well as possible adjustments arising from changes in detector cooling and installation requirements. In the industrial context of computational fluid dynamics, Wall-Modeled Large Eddy Simulations (WMLES) and hybrid Reynolds-Averaged Navier Stokes (RANS)/LES models suffer from multiple shortcomings. The present thesis tackles this problem, and a Generalized WMLES (GWMLES) model is presented in the fourth chapter that enables the modeling of the entire log-layer. GWMLES is validated against models of different fidelities and experimental data. The model developed gives a level of accuracy similar to recent LES results with a much lower computational cost. It is demonstrated that it is more accurate than URANS even when the resolved portion of the energy spectrum is reduced, and that it requires up to half the computational cost compared to the Detached Eddy Simulation (DES) model.