The management of toxic leaks in confined environments, such as CERN's experimental caverns and space habitats, is crucial due to potential risks to human life. This thesis examines aggressive carbon dioxide releases from CERN's 2-Phase Accumulator Controlled Loop (2PACL) system, used in both terrestrial (ATLAS, CMS, LHCb) and extraterrestrial (AMS) experiments. Computational fluid dynamics (CFD) simulations were performed to assess the severity of various failure modes. A two-dimensional analysis of the immediate leak area showed that shock waves travel up to 30 centimeters from the failure point. Subsequent three-dimensional simulations in the CMS experimental cavern (UXC55) revealed that while the upper cavern floors are well-protected, the lower floors pose significant risks due to semi-confined spaces and worse ventilation. Recommendations include the use of oxygen deficiency monitors, self-rescue masks, and detailed evacuation plans to enhance worker safety.

Analysing and developing turbomachinery at off-design conditions requires the usage of robust and efficient Computational Fluid Dynamics (CFD) solvers. With the introduction of the computer, many advancements have been made in the field of numerical methods involving these flow problems. These numerical methods are used in order to solve these complex flow structures that are formed due to the interaction between the rotating machinery and the fluid. With the emergence of new design paradigms using computer-aided optimisation, novel solver methods are required which are able to deal with the increase in computational cost and the convergence difficulties following off-design conditions. One of the CFD solvers that is used for turbomachinery design is the open-source software SU2. SU2 is currently developed partially at the Delft University of Technology, where an increase in solver performance with respect to turbomachinery aerodynamics is greatly desired. The current work provides a numerical study of SU2's current performance with respect to turbomachinery analysis, as this is currently unknown.

The current performance of SU2 is to be analysed using steady RANS turbomachinery simulations. The research conducted by Xu et al. \cite{Xu2020} will be used as reference data in order to validate SU2's performance together with data obtained from the CFD solvers CFX and Numeca. The research conducted by Xu et al. resulted in the development of their NUTSCFD solver, which showed strong performance with respect to turbomachinery analysis. Four test cases are set up and used in order to analyse SU2. The test cases that are considered for the numerical study include: the NACA 0012 airfoil, the LS89 turbine cascade, the MTU centrifugal compressor and the 1.5 stage ETH turbine. The first three test cases are validated using the results obtained by the NUTSCFD solver, where the ETH turbine is validated using CFX and Numeca. Following these test cases, SU2's performance is analysed using the residual behaviour. Xu et al. dedicate their performance increase to be the result of the Newton-Krylov method that is implemented in their NUTSCFD solver. SU2 also includes a Newton-Krylov solver, but its performance with respect to turbomachinery is also unknown. A performance assessment with respect to SU2's Newton-Krylov method involving turbomachinery analysis is therefore conducted as well.

The results obtained using the NACA 0012 and LS89 test cases show a discrepancy in solver performance involving SU2. With respect to SU2's Newton-Krylov solver, the NACA 0012 test case shows a reduction in non-linear iterations where this is not found for the LS89 test case. Both results showed however a large difference in performance when compared to the NUTSCFD solver, where SU2's standard solver was also unable to match NUTSCFD. The results obtained following these test cases have led to the development of the MTU test case, where the MTU test case was to be used in order to provide a more accurate comparison between SU2 and NUTSCFD. Instead, SU2 was unable to run the MTU test case, where the solver showed stalling behaviour...

Studies revolving around data-driven methods have been on a rise in recent years to improve highly modelled methods such as the two-equation turbulence models of Reynolds-averaged Navier-Stokes (RANS). Similarly, such data-driven methods are implemented into partially-averaged Navier-Stokes (PANS). PANS is a young bridging method that fulfils the requirements of bridging methods set by Speziale [1]. It can be adjusted according to the fraction of a flow field that is desired by the user to be resolved and modelled by changing the value of fk, a parameter that takes a value between 0 and 1. In this thesis, PANS is extended via a combination of two data-driven methods: k−corrective frozen RANS [2] and data-driven stochastic closure simulation (DSCS) [3]. The k−corrective frozen RANS method aims to correct for the model errors in the k−equation and the anisotropy of the Reynolds
stress tensor derived from the Boussinesq approximation. While DSCS also aims to correct for the anisotropy of the Reynolds stress tensor, it considers that PANS solves for the unresolved part of the flow field and thus corrects for the unresolved anisotropy. While PANS and the two data-driven methods have independently been proved to work as they were theoretically desired, combining these ideas has not yet been attempted. As any turbulence kinetic energy solving RANS turbulence model can be developed into PANS form, the k − ω SST model was chosen for the best initial prediction. This SST model for PANS is extensively derived and then reformulated to produce the target correction terms. The correction terms are analysed at various values of fk and they show good agreements with fk.

Flat surfaces with arrays of wall-normal pores called micro-cavity arrays have shown potential to reduce turbulent skin friction drag. Their designs have been inspired by acoustic liners which are sandwich panels consisting of a honeycomb core, a perforated top plate and a solid back sheet used on aircraft engine nacelles for noise abatement. Research on acoustic liners has paved the way to identify the parameters that are crucial for optimising the design of micro-cavity arrays. Spanwise rectangular grooves, a special case of cavities, have also shown potential to reduce skin friction drag. This thesis aims to validate the reported drag performance of micro-cavity arrays and grooves through experimental methods used in literature and direct force measurements. 

In this research, experimental results show that although pores and grooves are capable of reducing turbulence energy in the boundary layer, there is a significant drag increase with respect to a smooth reference. Based on measured drag performance and past research on cavity flows, it has been argued that the observed drag increase is caused by pressure forces acting on the inner walls of the pores and grooves. As past research on micro-cavity arrays did not perform direct force measurements, the contribution of pressure drag had not been investigated. Pressure drag is expected to be present and perhaps even dominate over skin friction drag reductions. Further research into the flow mechanics inside the pores is required to validate this argument and find a way to minimise the pressure forces. Wall-normal pores could still be an effective turbulent drag reduction technique if arrays are optimised for skin friction drag reductions that surpass pressure drag.