Investigations of Scaling Effects on Submarine Propeller and Stern Boundary Layer Flows

摘要

A common problem in submarine design is the inability to build and test models at conditions that match the Reynolds number of prototype submarines, leaving computational simulations as the only method for testing new designs. A study was performed in which experimental and computational estimates of drag and propeller performance were obtained using models which differed in length by an order of magnitude. The purpose was to see how well the computational results would track the experimental results over the large jump in length, so that insight might be obtained into how well the computational results would predict the full-scale flows at lengths an order of magnitude larger than the large model used. The computations predicted the propeller performance well over the range of parameters simulated. The computational simulations under-predicted the drag of the bare submarine hull at the largest Reynolds number tested for the large model. This difference was attributed to limitations in the grid and/or turbulence model, and showed that significant care would need to be exercised in extending the computations to very large Reynolds number.