Ships underwater radiated noise is a subject of great and increasing interest in naval architecture because of its impacts on the environment and on the on-board comfort. Among the different noise sources, the propeller is usually the dominant one, especially when it cavitates. For this reason a lot of efforts have been spent in studying the cavitation noise spectra in order to be able to predict its main characteristics at design stage. Nowadays, in order to reach this goal, the state-of-the-art solution is to perform model scale experiments. Unfortunately, this requires time-consuming tests in a cavitation tunnel with a model of the propeller and results are affected by scale effects that must be correctly interpreted. For these reasons, in this paper authors propose a hybrid approach which may be adopted to predict the main characteristics of the cavitation noise spectra without requiring an actual experiment. Moreover, the same approach may be used also in order to overcome typical model scale problems. For this purpose a hybrid modelling approach able to exploit both the physical knowledge of the problem and the real data obtained from many cavitation tunnel experiments performed on different propellers in different working conditions have been developed. Results on real data in both interpolation and extrapolation tasks will support the validity and the effectiveness of the proposal.
|Titolo:||Model scale cavitation noise spectra prediction: Combining physical knowledge with data science|
|Data di pubblicazione:||2019|
|Appare nelle tipologie:||01.01 - Articolo su rivista|