Kwon, SuneunSuneunKwonHur, JaewookJaewookHurPark, JeongyongJeongyongPark2024-04-122024-04-122024-04-048th International Symposium on Marine Propulsors (smp 2024)978-82-691120-5-4https://hdl.handle.net/11420/46471To enhance the fuel efficiency of large commercial vessels, various types of Energy Saving Devices (ESDs) are employed. Among these devices, the duct-type ESD with fins is the most widely used device in shipyards. However, it is crucial to prioritize the consideration of cavitation performance during the design stage of this ESD, as it significantly impacts propeller cavitation. In this study, the open-source program OpenFOAM was used to investigate the behavior of cavitation, fluctuating pressure and erosion index according to the design variables of the duct-type ESD, called Hi-PSD® (HD Hyundai Pre-Swirl Duct), for large container carriers. The resulting database obtained from numerical analysis using OpenFOAM allowed the analysis of correlations between these factors and design variables. This database was then employed to train a deep learning algorithm capable of predicting the effects on fluctuating pressure and erosion index induced by the change of Hi-PSD® geometry. Consequently, this trained deep learning algorithm can be utilized to predict fluctuating pressure and erosion index based on Hi-PSD® design parameters. Through this research, a comprehensive design process of Hi-PSD® for large commercial vessels is established for predicting cavitation performance, encompassing fluctuating pressure and erosion risk index.enhttp://rightsstatements.org/vocab/InC/1.0/ESDMarin PropellerCavitationHull pressure fluctuationOpenFOAMEngineering and Applied OperationsConference Paper10.15480/882.933410.15480/882.933410.15480/882.9294Conference Paper