digilib@itb.ac.id +62 812 2508 8800

The technological advances in underwater vehicles have increased the demand for the underwater thrusters. As propulsion system, the underwater thrusters are widely used for both manned and unmanned vehicles. The highest hydrodynamic performance fulfilling the market needs becomes a challenge in the development of thruster. This research addresses the design of underwater thruster based on reverse engineering method and hydrodynamics performance improvement using multi-objective evolutionary algorithm. A detailed thruster breakdown, components function analysis, and 3D model reconstruction of thruster propeller obtained from 3D scan data are presented. In addition, a structural analysis due to the operating depth and hydrodynamic simulation using high fidelity software are also included. The results of the hydrodynamic simulation are validated by experimental data. Furthermore, the optimization of thruster performance is only focused on propeller geometry improvement. The objectives of optimization are to maximize the thrust and minimize the input power. The current propeller geometry is parameterized into several geometry variables including root and tip blade angle, blade number, rake distance, and chord distribution. Those variables may give significant or less significant effects to the objectives of the optimization. Therefore, in order to reduce the design spaces and only involving the critical input variables, a sensitivity analysis is conducted. The remaining variables are then used in the optimization process. The algorithm used is the Strength Pareto Evolutionary Algorithm 2 (SPEA2) which is currently used in many practical design problems. The results of optimized geometry shows that there is 2% increment of thrust and 7% decrement of input power.