Crossflow transition modelling for a marine propeller at model scale

Paper i proceeding, 2023

One of the options to obtain the performance of a propeller at full scale is to extrapolate from data obtained from model scale testing. The Reynolds number of the model scale experiments typically leads to non-negligible extent of both laminar and turbulent flow, unless the flow is tripped, having a significant impact on the propeller performance. When using Computational Fluid Dynamics (CFD) methods to simulate the flow at model scale, the modelling of laminar to turbulent transition and its most important triggering mechanisms is critical in order to achieve accurate results.

The transition of laminar to turbulent flow is due to the instability of the laminar boundary layer with regards to disturbances that are present in the flow. Depending on the amplitude of these perturbations, different transition mechanisms take place, such as natural or bypass transition. In addition,
flow separation can also lead to transition. While the previous effects all can be considered as existing in two-dimensional flows, some only occur in three-dimensional flows, like crossflow transition. This mechanism occurs due to the instability of the crossflow velocity profile and is generally important in
surfaces with pronounced curvature, with such a case being propeller blades.

Well established turbulence models used with the Reynolds-averaged Navier-Stokes (RANS) equations, e.g. the Spalart-Allmaras or k − ω models, fail at accurately predicting transition, since their calibration is done for high Reynolds numbers flows. These models lead to a negligible extent of laminar flow, even at moderate Reynolds numbers of around 105 − 106. To accurately predict transition in RANS simulations, transition models are required. The most used transition models in CFD nowadays, such as
the γ − Reθ and γ models, handle some of the aforementioned transition mechanisms such as natural, by-pass and separation-induced transition. However, their original formulation was not designed to account for crossflow transition, and is unable to predict transition occuring due to this phenomena. Nonetheless, several extensions to the models have been proposed in the literature to address this shortcoming.

The goal of this work is to demonstrate the importance of crossflow transition modelling for a model scale marine propeller and how it impacts the propeller performance across a range of varying advance coefficients. Simulations for a controllable pitch propeller using the RANS equations with the k−ω Shear Stress Transport (SST) turbulence model and the γ transition model will be performed. The results from using two alternative crossflow extensions are compared in terms of propeller performance and flow over the blade surface. Baseline computations without crossflow modelling are included as well.