Experimental study of the effect of forward speed and following waves on roll damping of fishing vessels

Abstract

An extensive experimental program has been carried out to estimate roll damping parameters for three models of fishing vessels having different hull shapes and moving with forward speed. Roll damping parameters are determined using a novel method. This method combines the Energy method and the Modulating Function technique. The results show that this method gets better estimates compared with the original Energy method. -- A data processing system was designed to process the experimental data. A parameter Cerror was introduced to measure the error in roll damping identification. A database system was developed using VAX-Pascal to store the analytical results and perform various kinds of analyses. The data management and processing system in this research work has proved to be very efficient. -- The effect of forward speed, initial angle and natural frequency on roll damping is discussed. The effect of forward speed on roll damping was found to be nonlinear. The effect of initial angle is strong at zero and low forward speeds and decreases as the forward speed is increased. The effect of natural frequency was found to be weak. -- Ikeda's method was used to predict the roll damping coefficient. The results were compared with the experimental data. It was found that Ikeda's method overestimates the roll damping at higher forward speeds for all three models. This method fails in predicting the eddy damping for ship forms with hard chines. It was noticed that as models move with forward speed, their mean drafts increase. A modification to Ikeda's formula is proposed, making use of this observation. The values predicted by the modified formula fit the experimental data very well. -- A preliminary experiment has been done to investigate the effect of following waves on roll damping. It has been found that estimating roll damping parameters, without allowing for the time variation in the restoring moment, results in overestimating the values of these parameters. Further work is needed in this area.