Optimization of propeller pitch and revolutions in behind condition

Abstract

To reduce costs and limit pollution from transportation of goods by sea, a method for optimization of the propulsive efficiency of an 8000 dead weight tonnage tanker is investigated. Emphasis is put on the interaction between the propulsion system and the hull.The optimum combination of pitch and propeller rate of revolutions is sought for a controllable pitch propeller, working behind a tanker. Two stern shapes have been examined at the vessel s design speed, and in bollard pull condition. The optimization study was carried out on the hull with lowest resistance. All investigations are done in model scale.The optimization is done both numerically and experimentally. In the software SHIPFLOW, a zonal approach is used to simulate the flow around the ship. That is, the flow regime is divided into a potential-, a boundary layer- and a complete viscous-region. A lifting line propeller model is iteratively coupled to the viscous flow solver, for the self-propulsion simulation.Resistance-, propeller open water- and propulsion- tests are carried through in MARINTEKs towing tank, in Trondheim. The resistance test for one stern shape was repeated four times to determine the uncertainty in the measurements. All the tests were done at design draught in calm water. Five propeller pitch settings was tested around the design pitch, at three propeller loadings.The experimental results in terms of power, shows some degree of scatter. The lack of repeated tests for the propulsion case makes the uncertainty unquantifiable. There seems to be a trend that the higher pitch ratios perform better than the low pitch settings, for all propeller loadings tested.From the SHIPFLOW simulations it is found that the pitch/diameter=0.77 (P/D=0.77) (design P/D=0.87) performs best, by 0.5% better than P/D=0.67. It is seen that the thrust deduction increases with increasing P/D. This opposes the trends from the experiment. The maximum number of iterations was reached for the viscous flow solver, so the solution is not fully converged.The validation of the simulated results against the experiment is inconclusive. To limit and quantify the uncertainty, repeated propulsion tests is recommended.The simulation method used in this thesis looks promising, but further verification is needed.An interesting continuation of the work done in this project, would be to do the simulations in full scale, with a corresponding full scale validation trial.