Study on a procedure of structural safety assessment for an energy saving device subjected to hydrodynamic force

Abstract

Due to the soaring oil price and the demand of CO2 reduction related to environmental issues, the demand for the reduction of fuel oil consumption is greater than ever before. In this respect, various types of energy saving devices (ESD) have been developed. ESD is a kind of fin placed along streamline and installed around propeller or stern to improve the propulsive performance. The main direction of hydrodynamic force on the fin is nearly same as the streamline, and its magnitude may be negligible in calm sea. However, in harsh environment, the heave and pitch motion of a vessel becomes larger and the fin-shaped ESD would experience large out-of-plane load and there is a high risk of structural failure and fatigue damage. In a conventional design approach, Morisons equation may be adopted with constant coefficient for hydrodynamic force evaluation. Spectral approach has been also widely used based on the assumption of linear system. However, it is difficult for Morisons equation and spectral method to estimate hydrodynamic force exactly. Therefore, this study proposes a new procedure of structural safety assessment for energy saving devices (ESD) subject to hydrodynamic force and applies the proposed procedure to the fin-type energy saving device. The proposed safety assessment procedure consists of three main parts, seakeeping analysis, computational fluid dynamics (CFD) analysis and long-term analysis. As the sea-keeping analysis, potential based commercial code, WASIM, is used. Response amplitude operators (RAO) and response spectrums of vertical velocity at ESD are calculated. In CFD analysis, Hydrodynamic force are calculated for predefined regular waves using VOF (Volume of Fluid) and DFBI (Dynamic Fluid Body Interaction) techniques and a neural network is trained using the data. Irregular time histories of vertical velocities are generated from response spectrums obtained from sea-keeping analysis. In order to take into account the randomness of the irregularity, twenty different irregular time histories are generated. Then, each time history of vertical velocity is converted to time histories of hydrodynamic force. For each sea state, twenty maximum hydrodynamic force values for 3 hours duration are collected and Gumbel distribution is used to fit the data. This process is repeated for all sea states in wave scatter diagram and long-term value is calculated. An approximate long-term calculation is made using contribution coefficient based method. The method enables to carry out time domain analysis for a part of sea states that have dominant contributions to long-term exceedance probability. The contribution coefficients of all sea states can be calculated from frequency domain with less computational time. As a result, the total computation time for long-term analysis is reduced. Additionally, a procedure of fatigue strength assessment is established. 3 hours time series of vertical velocity is generated from the response spectrum and the peak values of vertical velocity are transferred to lift force and moment using the trained neural network. The time series of lift force and moment are transferred to stress histogram using a stress response per unit force. Finally, fatigue damage is calculated using Miners rule.