Analysis of ship maneuvering difficulties under severe weather based on onboard measurements and realistic simulation of ocean environment

Highlights

• Problems to maintain course in actual sailing are investigated.

• Data from spatiotemporally varying environmental fields are collected.

• Measured data are utilized to reproduce realistic fields.

• Rudder attenuation under severe weather is found by a statistical analysis.

Abstract

The increasing demand for efficient, safe, and economic operation of ships has drawn attention to practical maneuvering behaviors for developing autonomous ships. Actual sailing conditions are reproduced to determine the relationship between environmental factors and ship steering records in rough seas. First, we generate realistic ocean environmental fields and analyze actual sea data. Then, we derive a modular maneuvering model reflecting environmental disturbances for further simulations. The correlation and multi regression analyses are performed based on measured data and environmental factors, which illustrate that the abnormal rudder angles are caused by reduced steering effectiveness. Finally, an attenuation function acting on the rudder normal forces is proposed to simulate this type of reduction. The time histories of maneuvering difficulties are selected as verification datasets. The maneuvers are simulated by adopting the attenuation function, and the simulation results show fair agreement with the measured data. The significant wave height, wind speed, mean wave period, current speed, wind apparent direction, and wave encounter angles are found to be the most statistically significant factors of rudder attenuation in the studied cases. The results and conclusions obtained from this study are of great significance for the further exploration of actual ship maneuvering behaviors in seas.

Introduction

With the development of the shipping industry, shipbuilding companies and related research institutions are pursuing ship autonomy. The concept of maritime autonomous surface ship (MASS) was introduced at the 98th Session of the Maritime Safety Committee (2017) of the International Maritime Organization. In addition, this organization has recommended conducting theoretical and experimental studies related to MASS and developing different levels of autonomy. There are four nonhierarchical degrees of autonomy for a MASS according to the Maritime Safety Committee (2018): 1) automated processes and decision support, 2) remote control with seafarers on board, 3) remote control without seafarers on board, and 4) full autonomy.

In the last three years, pioneering research has been conducted to develop different degrees of MASS autonomy considering favorable fundamental conditions in near-shore areas (Reddy et al., 2019). These conditions include unimpeded network communications and a stable meteorological environment. However, ensuring safe MASS navigation in the open sea remains challenging due to the complex and changing environment. In addition, ship route design is restricted by meteorological conditions and the environment, which may increase fuel consumption and costs (Vettor and Guedes Soares, 2016). Moreover, load shifting risks, maneuvering difficulties, and capsize may occur under disturbances caused by severe weather (Sahoo et al., 2019). These potential risks greatly hinder navigation safety and may undermine the economic benefits of using a MASS. Therefore, the operational problems during actual sailing should be analyzed, especially regarding rare events occurring under severe weather. This analysis may reveal relations between the environment and ship navigation, thereby fostering the development of MASS.

The influence of the ocean environment on ship navigation has remained a research hotspot in marine engineering over time (ITTC Maneuvering ITTC Manoeuvring Committee, 2017). Although the maneuverability of ships under adverse weather has not been explicitly included in regulations, it became necessary since the introduction of corresponding guidelines in the EEDI regulations (International Maritime Organization, 2018). Several studies have addressed the characteristics of ship navigation under different environmental disturbances. Ueno et al. (2017) conducted tank tests to investigate different rudder and propeller control methods. However, they recognize that measurements in real scale may allow describing highly nonlinear maneuverability under adverse weather. Ruiz et al. (2019) investigated maneuvering considering still water and regular waves through model tests. They analyzed the effects of waves on the rudder and propeller based on experimental data. The Maneuvering Modeling Group (MMG) model is widely used to simulate 3-DOF and 4-DOF maneuvers under calm water or wind-wave conditions (Yasukawa and Yoshimura, 2015). The hydrodynamic performances of propeller and rudder during maneuvers are analyzed by Guo et al. (2018) based on the MMG model and CFD method. The propeller side force and asymmetric flow-straightening effect are taken into account in their study. Sukas et al. (2019) developed a feasible code to investigate the maneuverability of any low-speed ship with single-rudder/single-propeller or twin-rudder/twin-propeller configurations based on the MMG model. Besides, various methods could be applied to obtain the parameters in the MMG model. Liu et al. (2017) assess the existing empirical methods relevant to the MMG models' parameters, and an integrated maneuvering model for inland vessels has been developed using suitable empirical methods and RANS results. The viscous CFD method with the overset grid is applied to identify all the necessary parameters for the MMG model by Sakamoto et al. (2019). A method was proposed by Yasukawa et al. (2019a) for predicting steady sailing conditions under environmental disturbances, which effectively captures the maneuvering limit of ships under wind and wave disturbances. For instance, a ship may lose its steering ability and drift when sailing in irregular beam waves. Inspired by the work of the former, Jing et al., 2020a, Jing et al., 2020b proposed a practical method to construct a hydrodynamic coefficients database based on the stereolithographic model to improve the maneuvering simulation with the constantly changing wave conditions. Acanfora and Rizzuto (2019) investigated nonlinear effects when predicting motions of a ship in beam seas at zero speed and performed a comparative simulation analysis between different models considering idealized wave spectra. Paravisi et al. (2019) developed a simulation environment to test control strategies of unmanned surface vehicles. In the simulations, wind and current fields were integrated to compute the overall environmental effect. Aung and Umeda (2020) performed various maneuvering simulations considering not only the wind and waves but also the engine load limits of a ship. The initial values in the time-domain simulations showed notable effects on the ship trajectories, but the conditions after reaching steady state converged. Moreover, the ship trajectory was found to be more important than the equilibrium speed to determine the safety of a ship sailing in adverse conditions. However, these simulations were based on ideal conditions, such as constant scale and direction of winds and waves, and the ocean current was not included. Analyzing maneuvering behaviors in real situations might reflect unaccounted sailing effects. For instance, Tang et al. (2020) integrated steering and the effects of non-uniform flow fields obtained from a numerical model to investigate the maneuvering performance in restricted waters (e.g., inland rivers).

Besides simulations, various onboard measurement systems have been developed in recent years to evaluate ship performance in actual sea. In addition, sea trial data or onboard measured data have been used to investigate ship performance. Chen et al. (2015) analyzed the performance of ships under ocean currents and generated high-resolution Kuroshio currents from a numerical model. Lu et al. (2017) estimated waves and winds for rough-sea sailing in the Southern Hemisphere and compared the results with measurements from a 28,000-DWT class bulk carrier. Tsujimoto and Orihara (2018; 2019) thoroughly reviewed performance prediction methods and validation results of full-scale ships in the sea. You et al. (2020) estimated the actual sea margin of a liquefied natural-gas carrier using maneuvering equations. Although realistic winds and waves were obtained from the European Center for Medium-Range Weather Forecasts, the environmental conditions were fixed according to the time and position of a simulated ship.

Overall, most available methods to evaluate ship maneuvering and performance are based on model tests in water basins and numerical simulations. Research on maneuvering in actual sea is scarce, despite model ships failing to reflect full-scale ship maneuverability. Moreover, generating winds, currents, and waves for full-scaled ships resembling the actual sea is difficult. Thus, a fixed maneuver scheme, such as turning, zigzag, and emergency maneuvering, is adopted in many numerical simulations (Jing et al., 2020a, Jing et al., 2020b). In practice, however, the rudder angle is controlled by a human operator in a discrete form. On the other hand, wind is assumed to be steady in simulations, and short-crested irregular waves are commonly generated considering idealized wave spectra and spreading functions. These simplifications are adopted because the exact directions of wind and waves are difficult to estimate, especially when they are not aligned. Consequently, actual maneuvering may substantially differ from the ideal simulation results. In actual sailing, ships encounter following seas or bow/stern quartering seas. However, most physical experiments and numerical simulations are limited to head seas. Furthermore, few studies have considered the surface current effect, whereas most studies have been focused on wind and waves, both regular and irregular. Consequently, actual sailing cannot be analyzed comprehensively due to insufficient reliable sensor data under realistic environmental conditions, especially when sailing under severe weather.

Sasa et al. (2015) conducted long-term onboard measurements from a 28,000-DWT class bulk carrier from 2010 for the optimization of ship routing. Despite mechanical problems that impeded measurements in some periods, the measured data include waves, ship motion, navigation, and engine parameters during many severe weather events in both the Southern and Northern Hemispheres. As speed loss is the key indicator for optimal ship routing, it was thoroughly investigated using data acquired in rough seas (Sasa et al., 2017). From these studies, the authors found that maneuvering in rough sea differs from that in calm sea. For instance, according to its deck log, a ship encountered remarkable maneuvering difficulties in June 2013 due to severe weather, which caused problems to maintain the ship course. The corresponding onboard measurements show that the ship was frequently steered with large rudder angles during this period. However, the reasons underlying this maneuvering behavior remain unclear.

In this study, we investigated the abnormal maneuvering of a ship under severe weather. First, the onboard measured data are analyzed. Then, the environmental fields and maneuvering motions were reproduced based on numerical models and measured data. Further, a rudder attenuation function is proposed based on a statistical analysis of simulation results. More simulations are performed with the attenuation function. The results were compared with the measured data, which reveal the relations between environmental factors and maneuvering motion. Following this introduction, section 2 describes the onboard measurements used in this study. Section 3 details the methods to reproduce ocean environment and ship maneuvering in simulations. Section 4 reports the results of the reproduced winds, currents, waves, and maneuvering behaviors. In addition, results from statistical analyses based on the reproduced environment and measured data are presented. Finally, we summarize the major findings from this study in section 5.