An improved panel method for the prediction of performance of ducted propellers
MetadataShow full item record
An improved perturbation potential based lower order panel method is applied to the three dimensional problems of flow around ducts and ducted propellers. One significant development of this method is the application of full wake alignment scheme in which the trailing vortex wake sheets of the blades are aligned with the local flow velocity by also considering the effects of duct and duct wake. A process of repaneling the duct is simultaneously introduced to improve the accuracy of evaluation of the method. The results from the improved wake model are compared with those from a simplified wake alignment scheme (PSF-2 type scheme). At the same time, full-blown RANS simulations of the three dimensional problem are conducted. The forces, i.e. thrust and torque on the propeller predicted by the present panel method under the improved wake alignment model show good agreement both with experimental measurements, a hybrid method developed by the Ocean Engineering Group of UT Austin, and the full blown RANS simulations. Moreover, detailed pressure distribution on the blades and duct are compared among the various methods. The interactive method which couples the lower order panel method with a two dimensional boundary layer solver through a wall transpiration model is also introduced. An assumption of two dimensional flow is made on the individual stations of a three dimensional geometries and the coupling is applied in a stripwise manner by including the interaction effects from other strips. The interactive method is then validated through the cases of bare ducts and ducted propellers. An important improvement is also made on the extension scheme for hydrofoils and propeller blades with blunt trailing edge. A more physical criterion has been established for determining the location of cut planes or the starting points of the extension. The extension scheme is applied to both axisymmetric problems and fully three dimensional problems. Correlations of results with experimental measurements and RANS simulations are presented.