Based on the RANS method, numerical simulation of self-propulsion model test is performed for the KVLCC2 model combined with Proportional-integral-derivative (PID) control theory. The initial propeller revolutions are estimated according to the numerically predicted ship resistance and propeller open-water performance, and sliding mesh technology is adopted to simulate the propeller operated behind the ship hull. Using the constructed Proportional-integral (PI) controller, the model self-propulsion point is determined by adjusting the propeller revolutions according to the total force acting on the ship model. The numerical results agree well with the published experiment data, which verifies the effectiveness of the method proposed in the present study.
The self-propulsion experiments are the primary approach to evaluate the ship resistance and performance, including resistance test, open-water propeller test and self-propulsion test. The traditional self-propulsion tests are performed for a model captive to the carriage in a towing tank. For a given target speed, runs at different propeller revolutions are performed and the net towing force is measured. With the net towing force, the propeller revolutions that zeroes the towing force is obtained, indicating that the self-propulsion point has been achieved by balancing propeller thrust with total resistance.
The drawback of the traditional approach is the high cost of manufacturing ship models and conducting model tests, thus many researchers focus on conducting numerical simulation of self-propulsion experiments. With the development of Computational Fluid Dynamics (CFD) in recently years, the approach has been used in attempts to predict self-propulsion and powering characteristics with good success. In self-propulsion simulation, interaction among hull, propeller, and rudder is the main difficulty, because the flow field around stern is quite complicated. In addition, the temporal scale and spatial scale of the flow is different for hull and propeller. In the simulation of propeller rotation, the smaller time step and grid spacing is needed, which is computational resource consuming. For the disadvantage of direct simulation of propeller, many researchers try to simulate the propeller working behind the ship with body force model. Phillips et al. (2008, 2009) adopted a prescribed body force model to simulate the propeller rotation, while Choi et al. (2009), Wöckner et al. (2011), Rijpkema et al. (2013), and Starke and Bosschers (2012) used the Reynolds-averaged Navier-Stokes (RANS) solver coupled with an iterative body force model to simulate the interaction among hull, propeller and rudder. Although the accuracy is lower than direct RANS simulation, less computational resource is needed. The direct simulation of interaction among hull, propeller and rudder is also attainable at present. Tahara et al. (2006), Choi et al. (2010), Visonneau et al. (2012), Hayati et al. (2013), Bugalski and Hoffmann (2011), and Krasilnikov (2013) performed self-propulsion simulation with RANS method, considering the interaction among hull, propeller and rudder, with a discretized propeller represented with moving grid.