Computational Analysis of Submarine Propeller Hydrodynamics and Validation Against Experimental measurement
Seil, GJ and Widjaja, R and Anderson, B and Brandner, PA, Computational Analysis of Submarine Propeller Hydrodynamics and Validation Against Experimental measurement, Conference Proceeding UDT Pacific 2008, 4-6 November, Sydney, pp. CD Rom. (2008) [Non Refereed Conference Paper]
The hydrodynamic performance of a submarine propeller is critical to its operational capability. A well-designed
propeller provides optimum speed and endurance and with minimum vibration and acoustic emission. Within DSTO,
there exists an ongoing research program which addresses the modeling of submarine hydrodynamic performance using
both computational and experimental techniques. Together with the University of Tasmania (Australian Maritime
College) and industry, DSTO has been undertaking detailed open-water studies of submarine propellers. This paper
describes CFD calculations of a generic seven-bladed propeller which were performed over a range of advance numbers
using the commercially-available CFD codes: CFX 11.0 and FLUENT 6.3.26. The results of these calculations were
subsequently validated against propeller curves obtained from experimental measurements of the propeller made in the
Tom Fink cavitation tunnel at the Australian Maritime College.
The calculated thrust and torque were found to be in good agreement with the experimental measurements over the range
J=[0.5-0.9], with thrust and torque predicted by both codes being within 4% and 5% of the measurements, respectively.
The relative difference between the CFD calculations and experimental measurement increased for larger advance
numbers. The CFX results for thrust and torque were found to be in closest overall agreement with the experimental
measurements over the range of advance numbers calculated, whereas FLUENT provided the closest overall agreement
for open water efficiency. The propeller slipstream appeared to be well resolved close to the propeller where the mesh
provided good resolution of the flow field. However the slipstream circumferentially mixed out whilst retaining a
distinct radial variation. This may be attributed to a combination of mesh resolution and turbulence modelling. Future
research will provide experimental measurements with which to validate the slipstream.