Statistical analysis of a noise measurements for the series of roro ships

Abstract

The importance of noise and vibration measurements during the sea trials is significant, especially now when noise limits are getting more restrictive and ship-owners demand to obey those noise limits. For a new-built vessels the normal procedure is to perform noise forecast at design stage. Taking into account weight limiting the passive noise insulation becomes a very important issue, therefore in order to avoid solutions which are not effective one should base on a realistic noise excitation data. Noise forecasts are mostly based on initial documentation and characteristics of main sources e.g. engines, generator sets, propellers etc, which are provided (measured or calculated) by manufacturers. Ship noise prediction might differ from real on-board noise situation. The question is what is a standard deviation between forecast and noise measurement. Normally for noise prediction analysis the worst case scenario is taken into account, so one can expect that real on-board noise levels will be lower then forecasted. However, when weight limits are taken into account in standard work case scenario the real on-board noise levels are more or less the same as forecasted. The question is how much is it “more or less”? Conducting the measurements onboard of a series of ships creates the convenient opportunity to perform statistical analysis of the results. The paper focuses on statistical analysis of noise measurements performed for the series of ro-ro ships.