Analysis of CFD methodologies applied to a high-pressure H2 injector in support to the design process

This thesis presents a computational fluid dynamics (CFD) investigation of a high-pressure hydrogen injector developed in STAR-CCM+, with the aim of evaluating the capabilities, limitations, and computational efficiency of different numerical methodologies for predicting injector performance throughout the operating cycle. Particular attention is devoted to assessing their suitability as design-support tools capable of providing both quantitative predictions and physical insight into the complex flow phenomena governing the injection process. Two modeling strategies were considered: a steady-state (SS) approach based on discrete needle-lift configurations, and a Dynamic Fluid Body Interaction (DFBI) model employing an overset mesh methodology to model the needle motion from the balance of fluid and external forces. The steady-state methodology was first validated against experimental mass-flow-rate measurements and compared with an equivalent transient morphing model driven by the experimentally measured needle displacement. The objective was to assess the ability of the stationary approach to reproduce the experimentally observed injector behavior and to evaluate the extent to which steady-state simulations can capture the dominant physical phenomena predicted by a time-dependent formulation of the same problem. A simplified two-dimensional DFBI simulation campaign was subsequently carried out to further investigate the capability of a force-driven modeling approach to reproduce the cause-and-effect mechanisms governing injector dynamics and performance. An overset-mesh strategy was adopted to handle the needle motion, overcoming the limitations of the morphing technique in configurations involving contact or near-contact between moving and stationary walls. Mesh-sensitivity analyses were also performed to assess the influence of overset-grid resolution on solution accuracy, numerical stability, and computational cost. The results were then compared with those obtained from the steady-state methodology, highlighting the potential of stationary simulations as efficient and reliable tools for preliminary design studies. In contrast, the transient approaches proved more suitable for investigating complex unsteady and compressible-flow phenomena, providing deeper physical insight and making them particularly valuable for detailed design verification and final optimization.