Uncertainty quantification in the prediction of cavitation inception of the Duisburg Propeller Test Case

Accurate quantitative prediction of marine propeller performance and cavitaiton inception criteria has important implications for the eventual ship operator. However, there exist a wide range of uncertain parameters that need to be input to the CFD simulations and whose choice will affect the predictions made. As stated by Slotnick J. et al. (2014), the management of errors and uncertainties due to the lack of knowledge in parameters of a given fluidic problem is an important trend for studies involving computer simulations. Aiming to mitigate gaps in studies regarding the cavitation inception of marine propellers and to deepen the discussion on the subject, the present study seeks to quantify the uncertainties in the input data of a CFD simulation of the Duisburg propeller P1570. An existing uncertainty quantification (UQ) Python framework of Katsuno et al. (2021) capable of calculating the confidence interval (CI) and the Sobol indices (S) of each input variable is further developed and used to deduce the effect of physical input parameters on the simulation results.