Red-Emi: "Reduktion der hydroakustischen Emission von Propulsionssystemen"

Since the beginning of motorized shipping the anthropogene hydroacoustic noise level has risen continuously. Research has shown, that marine wildlife is negatively affected by this noise. Maritime regualtion societies have recently begun to regulate noise emissions from ships. In order to meet the current and coming regulations and to protect the biodiversity in the sea, the joint research project Red-Emi aims to evolve simulation techniques for the calculation of underwater radiation from ships.

During the design of propulsion systems the accurate prognosis of noise sources during operation can only be considered if efficient tools to calculate these are available. The approach in Red-Emi is two-fold: improving simulations and creating an online artificial neural network.

The simulation of noise radiated from cavitating propellers with propeller caps or rudders intersecting the propeller vortex sheet and cavitation will be improved by modeling the interaction with an efficient boundary element method. In addition to improving the methods for sheet and tip-vortex cavitation, a model for hub-vortex cavitation will be developed and the contribution of the gear in rudder propellers is investigated. Detailed RANS/LES simulations will be performed to verify the models.

In order to improve the efficiency in the early stages of the design process, an online artificial neural network (ANN) will be designed and trained successively. The successive training will provide a method to provide mutiple partners the possibility to share one ANN and profit from this network trained with a large data set. At the same time, the confidential training data is not disclosed.