Date of Award


Degree Type


Degree Name


Degree Program

Naval Architecture and Marine Engineering


Naval Architecture and Marine Engineering

Major Professor

Xiaochuan Yu

Second Advisor

Damon Smith

Third Advisor

Nikolas Xiros

Fourth Advisor

Brandon Taravella

Fifth Advisor

Linxiong Li


In marine and offshore engineering, dropped objects, such as drill pipes, anchor chains, containers and some small parts, can accidentally fall into the water from ships or offshore platforms, causing casualties on deck or damage to underwater equipment. Damaged equipment can further harm the environment, such as oil spills from damaged wellheads. Therefore, for safe engineering and environmental protection reasons, we need to develop methods and tools that can predict the trajectory of dropped objects.

In this dissertation, we first study containers dropped from ships. More and more containers are falling into the sea due to bad weather. Containers lost at sea can negatively impact shipping companies, traders and consumers, and the environment. The problem of locating and recycling discarded containers is a challenging engineering problem. We design and implement a series of model tests of small-scale container models to study their falling trajectories for retrieving. We first build a standard 20-foot container model in SOLIDWORKS. Then, export the three-dimensional (3D) geometric model in STL (Standard Tessellation Language) format to the Stratasys F170 Fused Deposition Modeling (FDM) printer. A total of six models, which are made of Acrylonitrile Styrene Acrylate (ASA), are printed for testing purposes. They represent three different loading conditions, different densities, and centers of gravity (COGs). The physical models were dropped into the towing tank of the University of New Orleans (UNO). It was found from the experimental tests that the effect of the initial position after sinking would lead to a certain initial rotational speed, which had a great influence on the lateral displacement, which in turn affected the final landing position. And it can be further observed that these models typically flip for approximately 0-5 cycles during a drop in a water depth of 1.8 meters. This series of model tests not only provided experimental data for the study of the trajectory of box-shaped objects, but also provided valuable references for offshore salvage operations and pipeline layout design.

Another focus of this dissertation is the numerical study of the trajectory of cylindrical objects. We first propose a state-space model of a dropped cylinder based on Aanesland's (1987) 2D equations of motion and Xiang at al. (2016)'s 3D equations of motion. Then we further investigate the heave-pitch coupling term in 3D theory, which was ignored in 2D theory, and find that this term significantly affects the trajectory of dropped cylindrical objects. Then in the state-space model, the original deterministic trajectory is described as a stochastic process by adding small perturbations that satisfy Gaussian distribution to the initial state of the dropped cylinder. Second, three probabilistic methods of state estimation, namely Monte Carlo (MC) method, unscented method, and Cubature method, are used to predict the trajectory envelope of dropped objects and provide reliable results. The MC method is a classical method for solving stochastic problems and has been applied to study the motion of dropped objects. However, it always requires a sufficiently large sample, resulting in a large amount of computation, which is not suitable for practical applications, especially real-time monitoring. The simulation results show that two other advanced statistical methods, unscented method, and Cubature method, give similar results compared to the MC method, but they consume much less computation time. Therefore, the first two methods significantly increase their application possibilities, providing an effective way for envelope prediction of dropped cylindrical objects in marine transportation or marine operations. Furthermore, the Cubature method requires no additional tuning compared to the unscented method, reducing the computation time, making it easier and more suitable for dynamic and real-time risk assessment of dropped objects for maritime transport and installation.


The University of New Orleans and its agents retain the non-exclusive license to archive and make accessible this dissertation or thesis in whole or in part in all forms of media, now or hereafter known. The author retains all other ownership rights to the copyright of the thesis or dissertation.