This thesis investigates the performance of vaneless radial and conical diffusers operating with supercritical carbon dioxide (SC-CO2) under various geometric configurations. The motivation arises from the growing interest in compact, high-efficiency turbomachinery suitable for supercritical fluid cycles. The study employs a combination of one-dimensional modelling (1D) and computational fluid dynamics (CFD) to analyse pressure recovery in view of frictional effects and flow separation. All the CFD simulations were conducted using commercial software Ansys CFX 2024 and 2025 integrated into Ansys Workbench and all the coding, including 1D modelling and post-processing was done with Matlab 2024. The performance of SC-CO2 operating diffusers was compared against air simulations at ambient conditions at matched Mach regime. The results show the advantageous pressure recovery of SC-CO2. Furthermore, the comparison between 1D and CFD modelling shows acceptable agreement for low expanding diffusers but reveals significant deviations in outcomes close to separated flow conditions.

While experimental studies typically ensure the proper scaling of the advance ratio, testing under ambient conditions often leads to a reduced free-stream Mach number and, if the advance ratio is maintained, to a reduced tip Mach number. This research analyzed the sensitivity of a propeller's far-field tonal noise emissions to variations in helical tip Mach number when operating at a non-zero angle of attack and experiencing wake encounter, using low-fidelity methods. The chosen aeroacoustic solver, based on Hanson's time-domain theory, allowed to capture physical insights via waveform analyses.
The analyses conducted showed that variations in tonal noise amplitude, directivity, and spectral content due to a flow non-uniformity depend on the propeller helical tip Mach number. An enhanced influence of steady noise sources at higher Mach numbers was observed. Waveform analyses suggest that, as Mach number increases, the effect of temporal variations in the acoustic planform—determined by blade element positions at their retarded times—on unsteady loading noise diminishes relative to steady loading noise, thereby amplifying the influence of the latter. These results highlight that propeller aeroacoustic data obtained at low helical tip Mach numbers in experiments cannot be directly scaled to full-scale propellers operating at high Mach numbers.

Computational fluid dynamics (CFD) is an important tool in design involving fluid flow. Scale-resolving CFD methods exist, but they are too computationally expensive for practical design. Instead, the relatively cheap Reynolds-averaged Navier-Stokes (RANS) approach is the industry standard, specifically models based on the Boussinesq hypothesis, which are unable to represent the effect of turbulence anisotropy. Development of RANS models based purely on physical arguments has stagnated; however, data-driven turbulence modeling presents a paradigm shift for improved predictions. Though this technique has produced successful models tailored to specific flows, it has yet to produce a successful general turbulence model, which is the focus of this work.

In this work, models consist of corrections to the classical k-omega SST turbulence model; bijDelta to correct the Reynolds stress anisotropy and R to correct the turbulent kinetic energy. High fidelity data combined with the k-corrective-frozen technique is used to obtain exact correction fields, which are validated. Then, the SpaRTA framework is used to regress symbolic expressions for the corrections. Using a newly developed solver, models are injected into a full RANS solver to assess a-posteriori performance for various test cases. SpaRTA identifies a good R model, but only after a-posteriori optimization of coefficients, this model holds promise for generalization. Meanwhile, SpaRTA is unable to find a good bijDelta model due to its reliance on linear regression. A new framework based on non-linear regression is introduced which identifies a much better bijDelta model, though this model is Reynolds number dependent.