Ship hull maintenance procedures are dictated by rules and guidelines set by IMO (International Maritime Organization), IACS (International Association of Classification Societies) and individual classification societies. These bodies impose a rigid schedule of planned surveys to which vessels must adhere. However, planned maintenance offers little to no assurance, in the intervals between surveys, of possible structural degradation (e.g. large cracks, excessive corrosion or buckling) which could lead to catastrophic failure (MOL comfort accident is such a case). Therefore, the need for developing methods/techniques that enable the condition-based monitoring (predictive maintenance) of ship hulls is highlighted. These can be found within the overarching framework of Industry 4.0, which among other entails the concept of the digital twin, that is regarded as a virtual representation of the ship’s hull structure [1]. Such an asset exploits data collected from onboard sensors, in real time, that are processed so as the hull’s operational condition is remotely monitored.The present work draws inspiration from these concepts and aims to take a first step towards this direction by proposing an optimization-based approach for solving the inverse problem, i.e. identification of a damage by processing response data. &nbsp;Idealized geometry and loading conditions for the deck and side shells are considered for assessing the method. &nbsp;A single hole (circular discontinuity that may be regarded as a nominal stress disturber) is considered as a damage randomly located within the defined domain. The optimization approach is assessed solely within a virtual environment. A Finite Element model that serves the purpose is developed to provide virtual strain readings, representing measured data from on-board sensors. The virtual sensors are positioned along the zero-strain path for each considered case. Axial strains at the ship’s neutral axis on the side shells and shear strains at the centerline of the deck. The idea is that once a damage occurs then strain readings will immediately be captured from the otherwise zero strain levels. &nbsp;&nbsp;The hole detection strategy amounts to predicting its, location, and diameter and these unknown properties are treated as design variables in an optimization problem. The way this is achieved is by minimizing an error function between strains measured for any given damage scenario and an indicative target scenario. The main objective of this work is to assess the effectiveness of three alternative yet well-established optimization algorithms, i.e. a gradient based method, a non-gradient heuristic based method (Genetic Algorithm) and a statistical based method (Design of Experiments and Response Surface Methodology).&nbsp;[1] Smogeli, O. (2016). Digital Twins at Work in Maritime and Energy. DNV GL.

An Optimization Based Approach for Damage Identification of Idealized Ship Structural Assemblies

Paper Proceedings: https://asmedigitalcollection.asme.org/OMAE/OMAE2021/volume/85116

OMAE 2021

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.