Developing Computational Methods for Detailed Assessment of Cavitation on Marine Propellers
Cavitation often brings negative effects, such as performance degradation, noise, vibration, and material damage, to a marine propulsion systems, but for optimum performance, cavitation is almost inevitable. Therefore, it is necessary to improve the understanding of cavitation in order to maximize the performance without encountering severe problems. Experimental tests can only provide limited information about this complex phenomenon. This thesis deals with improving numerical simulations methodologies that can offer a more complete picture of the cavitation process, making it possible to investigate the flow in more detail with some confidence, thus enabling an improved design.
Numerical simulations of non-cavitating and cavitating flows are conducted using OpenFOAM. The flow is modelled using Implicit Large Eddy Simulation and considering the two phases, i.e. vapour and liquid, as a homogeneous mixture through a volume fraction transport equation method along with the Schnerr-Sauer mass transfer model.
To avoid manual calibration of the mass transfer model coefficients, which may significantly affect both the accuracy and stability of the numerical predictions, an approach is suggested and tested to compute the mass transfer rate based on the flow local time scale during the solution procedure. Moreover, the saturation pressure is modified in order to take into account the shear stress effects on the liquid rupturing.
To test the proposed modifications, several test cases consisting of 2D and 3D hydrofoils and model scale propellers are simulated and the results are compared with experimental data. Integral quantities, local pressure data, and cavitation extent are studied for both the non-cavitating and the cavitating flows. Furthermore, the computational set-up is tested by varying domain size, mesh type and resolution, numerical schemes, and mass transfer model coefficients.
The overall results compare well with the available experimental data, provided the mesh resolution is sufficient. The proposed mass transfer model modifications give a considerably improved prediction of pressure distribution and cavity extent. Some results yield overpredicted cavitation, indicating discrepancies between the modelling approach and model scale experimental techniques.