Compressive strength of tubular members with localized pitting damage considering variation of corrosion features

Localized pitting corrosion often occurs on marine and offshore structures in the form of patch corrosion with great uncertainties in the location, size and shape. The variation of corrosion features affects ultimate strength of tubular members significantly, but it is still not well understood. This paper presents a numerical study on tubular members of diverse slenderness ratios to clarify the localized pitting effect on ultimate strength. Numerical analyses were performed based on novel models of pitted members that were calibrated against benchmark column tests. Corrosion pits were randomly introduced on the local outside surfaces of members via stochastic simulation, forming corrosion patches varied in the location, size and shape. Numerical results obtained were regressed to propose a unified empirical formula to predict ultimate strength. It turned out that the shape of the corrosion patch has a significant influence on the ultimate strength. The shape change of the patch can alter failure modes of medium length columns. The reduction of ultimate strength is closely related to the shape ratio of the patch besides the volume loss of corroded material. The unified empirical formula incorporating the shape ratio and the volume loss shows a good ability to predict the experimental results.     

Calibration of safety factors for offshore wind turbine support structures using fully coupled simulations

The offshore wind industry experienced a boost during the last decade in terms of size of wind farms and rated capacity of the wind turbines: towers are getting taller and blades are getting longer, constantly facing new and complex challenges. Because of the relative immaturity of the wind industry, and the fact that the offshore design standards stemmed from the oil and gas industry, it is generally acknowledged that the reliability levels achieved, although not very well understood, might result in partial safety factors not optimal for OWT. This paper addresses this situation by studying the reliability levels delivered by the current standards and assessing the validity of the safety factors through a reliability-based code calibration. The combination of the low probability of failure imposed on the design of OWTs and the computational cost of the aero-elastic time-domain simulations brings out the need to develop new approaches for reliability analyses. In this paper, the reliability analysis is performed using a Kriging surrogate model to approximate the load-effect from the aero-elastic simulations converting expensive-to-evaluate limit state functions to explicit functions. Subsequently, a calibration of the safety factors is carried out using the probabilistic models from literature. The approach is applied to an industry-reference turbine and support structure. The results showed very low probabilities of failure for the most severe design cases and confirm that the safety factors from the IEC are mostly adequate.     

Stochastic averaging for estimating a ship roll in random longitudinal or oblique waves

In this work, the C11 container ship is taken as an example to analyze its rolling performances in random longitudinal or oblique waves. Firstly, a dynamic model of C11 roll in random waves is improved, and it is verified by the model test and numerical simulation. Mathematically, this dynamic model is a one-dimensional stochastic differential equation with random parametric (and external) excitation. Secondly, an enhanced stochastic averaging method is proposed to solve this stochastic differential equation. The validity of the solutions was verified by Monte Carlo simulation. At last, the probabilistic characteristics of the extreme rolling response were investigated based on the calculated results using enhanced stochastic averaging method. According to the analysis, some advices for ship's manoeuvring can be put forward when ships are navigating in random waves.     

Uncertainties estimates of hull girder still water loads of bulk and dry cargo ships through Monte Carlo simulations

This paper deals with the estimate of uncertainties affecting still water hull girder loads of bulk and dry cargo ships. In strength assessment of ships, two main categories of acting loads are considered: still water loads and wave induced ones. While the latter are generally defined bearing in mind their stochastic nature, this is not the case for still water loads, which are basically deterministically considered. The underlying assumption is that there is an overall control of the operational profile during the service of a ship. However, this is not the case in actual fact, especially for general dry cargo ships and bulk carriers, since the loading/unloading process cannot be fully controlled by the crew, often resulting into loading conditions rather different from those planned by the designer. Based on an earlier work, where loading conditions of the above-mentioned ship types were statistically analyzed, in the present paper Monte Carlo simulations are used to estimate the uncertainties affecting the hull girder still water loads of ships in service, showing that their allowable values can be exceeded due to inaccuracies in ship and cargo management.     

Iterative methodology for response based analysis of an FPSO mooring system

Response based analysis (RBA) has been developed for prediction of extreme N-year return period responses and design metocean conditions of offshore structures. For applying the RBA, the behaviour of the offshore system subjected to a long history of metocean conditions needs to be predicted, and then, the probabilistic analysis is applied to estimate its long-term responses. Due to the large number of analysis cases required, the structural simulation is usually performed either by simplifying the structural model or by using computationally efficient tools, such as frequency-domain (FD) analysis. These approaches usually decrease the accuracy of predictions mainly when they are utilized for nonlinear systems. On the other hand, employing time-domain (TD) simulations leads to more accurate results but it is computationally expensive. Application of RBA for a weathervaning FPSO, which is the subject of the present study, makes TD analysis an essential requirement because of a highly nonlinear behaviour of the system. In the present study, an efficient methodology is proposed that aims at reducing the computational efforts of RBA by joint application of TD and FD simulations in combining the structural and statistical analyses through a single process, such that the number of time-consuming TD simulations is minimized. After initial screening using the results from FD simulations, the methodology identifies the response events (storms) that contribute the most to the N-year response and sets out an iterative process in which only those events that are most important are analysed by fully-coupled TD simulations. Within such events, a similar approach is also applied to intervals (sea states) where only the most contributing intervals are analysed in TD, and the remaining intervals are left for a less accurate FD analysis without sacrificing the overall accuracy. The proposed methodology provides a robust framework for distinguishing between “mild” and “severe” response events, without specifying any predefined limits for the metocean parameters or making a subjective judgement. Although it is developed for the mooring system of a weathervaning FPSO, it should also be applicable to any type of offshore structure and any structural response. This paper is the first part of the study and concentrates on the development of the efficient methodology to optimize the application of RBA to FPSO mooring systems, whilst its detailed application is subject of the second part of the study.     

Stochastic frontier analysis of excess access to mid-trip battery electric vehicle fast charging

This study aims to explore how factors including charging infrastructure and battery technology associate the way people currently charge their battery electric vehicles, as well as to explore whether good use of battery capacity can be encouraged. Using a stochastic frontier model applied to panel data obtained in a field trial on battery electric vehicle usage in Japan, the remaining charge when mid-trip fast charging begins is treated as a dependent variable. The estimation results obtained using four models, for commercial and private vehicles, respectively, on working and non-working days, show that remaining charge is associated with number of charging stations, familiarity with charging stations, usage of air-conditioning or heater, battery capacity, number of trips, Vehicle Miles of Travel, paid charging. However, the associated factors are not identical for the four models. In general, EVs with high-capacity batteries are initiated at higher remaining charge, and so are the mid-trip fast charging events in the latter period of this trial. The estimation results also show that there are great opportunities to encourage more efficient charging behavior. It appears that the stochastic frontier modeling method is an effective way to model the remaining charge at which fast-charging should be initiated, since it incorporates trip and vehicle characteristics into the estimation process to some extent.     

Scalable stability analysis on large connected vehicle systems subject to stochastic communication delays

In this paper large connected vehicle systems are analyzed where vehicles utilize vehicle-to-vehicle (V2V) communication to control their longitudinal motion. It is shown that packet drops in communication channels introduce stochastic delay variations in the feedback loops. Scalable methods are developed to evaluate stability and disturbance attenuation while utilizing the mean, second moment, and covariance dynamics in open chain and closed ring configurations. The stability results are summarized using stability diagrams in the plane of the control parameters while varying the packet delivery ratio and the number of vehicles. Also, the relationship between the stability of different configurations is characterized. The results emphasize the feasibility of V2V communication-based control in improving traffic flow.     

Study on mean-standard deviation shortest path problem in stochastic and time-dependent networks: A stochastic dominance based approach

This paper studies a mean-standard deviation shortest path model, also called travel time budget (TTB) model. A route’s TTB is defined as this route’s mean travel time plus a travel time margin, which is the route travel time’s standard deviation multiplied with a factor. The TTB model violates the Bellman’s Principle of Optimality (BPO), making it difficult to solve it in any large stochastic and time-dependent network. Moreover, it is found that if path travel time distributions are skewed, the conventional TTB model cannot reflect travelers’ heterogeneous risk-taking behavior in route choice. This paper proposes to use the upper or lower semi-standard deviation to replace the standard deviation in the conventional TTB model (the new models are called derived TTB models), because these derived TTB models can well capture such heterogeneous risk-taking behavior when the path travel time distributions are skewed. More importantly, this paper shows that the optimal solutions of these two derived TTB models must be non-dominated paths under some specific stochastic dominance (SD) rules. These finding opens the door to solve these derived TTB models efficiently in large stochastic and time-dependent networks. Numerical examples are presented to illustrate these findings.     

Modeling absolute and relative cost differences in stochastic user equilibrium problem

This paper aims to develop a hybrid closed-form route choice model and the corresponding stochastic user equilibrium (SUE) to alleviate the drawbacks of both Logit and Weibit models by simultaneously considering absolute cost difference and relative cost difference in travelers’ route choice decisions. The model development is based on an observation that the issues of absolute and relative cost differences are analogous to the negative exponential and power impedance functions of the trip distribution gravity model. Some theoretical properties of the hybrid model are also examined, such as the probability relationship among the three models, independence from irrelevant alternatives, and direct and indirect elasticities. To consider the congestion effect, we provide a unified modeling framework to formulate the Logit, Weibit and hybrid SUE models with the same entropy maximization objective but with different total cost constraint specifications representing the modelers’ knowledge of the system. With this, there are two ways to interpret the dual variable associated with the cost constraint: shadow price representing the marginal change in the entropy level to a marginal change in the total cost, and dispersion/shape parameter representing the travelers’ perceptions of travel costs. To further consider the route overlapping effect, a path-size factor is incorporated into the hybrid SUE model. Numerical examples are also provided to illustrate the capability of the hybrid model in handling both absolute and relative cost differences as well as the route overlapping problem in travelers’ route choice decisions.