On Methodology for a Digital Twin of Ship Propulsion Under Harsh Environmental Conditions

Abstract The propulsion system of a ship is a critical element for its safety and integrity. Therefore, it must be maintained properly to ensure that a vessel is able to fulfill its primary purpose. The need for maintenance is even more pronounced in ships where extreme environmental conditions, such as ice-covered waters, are expected. This is due to an increase in the magnitude and uncertainty of loads and therewith the increase in possible damages if the propulsion system fails. The use of digital twins is becoming increasingly popular in shipping and can also be used for intelligent maintenance and operation prediction of the propulsion system. This paper proposes the methodology of creating such a digital twin with a specific focus on the required measurement infrastructure, modeling of components, loads, and damage as well as how all of these aspects are combined. The digital twin was created using a polar supply and research vessel, S.A. Agulhas II. Bearings are discussed from the view of high-fidelity simulations and wear modeling. The propeller is modeled as a point mass with the propeller laws for predicting the hydrodynamic torque and thrust. Shaft dynamics are modeled using either a lumped-mass torsional model, a modal model, or a finite element model. Damage of the shaft and the propeller is based on SN-curve fatigue calculations. The motor torque is modeled using an equivalent circuit model while motor damage is modeled by estimating the hottest temperature within the motor windings. Ice-breaking simulations are used as inputs to propeller-ice interaction models to obtain insight into the loads caused by these interactions. The material properties of ice used during simulations are validated using experiments. The need for further validation for low-fidelity models and the need for reduced order modeling or surrogate models for high-fidelity models are discussed. Finally, the implementation of this digital twin is discussed based on the developed model and the found problems that still need to be overcome to increase the technology readiness level.