| Wave drag significantly affects ship powering and speed prediction calculations, and the performance of sailing yachts. Linear Neumann-Kelvin ship wave theory is taken as a starting point for the development of a computational approach for predicting water flow about yacht hulls. Previous work by Baar and Price, Newman, and Doctors and Beck for the calculation of the Kelvin source and its derivatives and their use with panel methods is repeated, refined and extended.; Consistent and accurate results are obtained for numerical tests and Farell's submerged spheroid test cases.; Baar and Price's results for the Wigley hullform (which compare well with experiments) were not able to be duplicated in the present study, despite significant tests of both local and integrated solution values. The tests did indicate that the current implementation was behaving correctly, and gave wave resistance results equivalent to those from Doctors and Beck's study, to a level of agreement which contrasted sharply with the wide scatter in Chen and Noblesse's survey. These results remove to some extent numerical inaccuracies as a postulated source of the theory's difficulties, leaving the conclusion that Neuman-Kelvin theory, as it is currently understood, does not give satisfactory wave resistance results for realistic ship hullforms.; These and further results lead to, and reinforce, the suggestion that the problem lies with the waterline integral term; a new treatment of this term may substantially increase the applicability of Neumann-Kelvin theory.; The extensions required for modelling sailing yachts are considered. Investigation of three yacht design problems show that the developed model can predict, and partly explain, previously observed free surface effects. These "relative" effects may be adequately predicted even though the absolute results appear less reliable.; For the simplified parabolic yacht hullform tested, the sideforce is relatively constant over the speed range generally relevant to upwind sailing, and well modelled by the (zero Froude number) cosine squared (heel angle) relationship. The effects of a bulb, and heel (for the Wigley hullform), on wave resistance are shown.; Suggestions for further work, calculation details, and tables of wave resistance and Kelvin source (and gradient) values are given. |
