Corrosion Prognosis: Maritime Structural Performances in Service Environments

For marine platforms, assessing the structural resilience in a corroded condition is vital for both design and maintenance practices. With the development of computational and experimental methods for structural analysis, the accuracy of the structural response prediction relies on a better understanding of the material degradation process. However, a realistic estimate of corrosion is inherently a complex undertaking. Corrosion of even a single form can often involve multiple stages, each of which has different steps across several geometric scales; corrosion systems are often multi-layered and involve geometric complexities; the mechanical factors (stress/strain distributions) could affect the corrosion initiation and kinetics. These complexities have resulted in scientific barriers to the advancement of a corrosion prognosis that forecasts damage accumulation, as well as a computational realization of the corrosion-structural analysis. This paper reviews the numerical and experimental work that the authors have done, including the development of nonlinear finite element models to assess the behavior of damaged steel ship structures, full-field experimental verifications, application of the mechanoelectrochemical theory and in situ tensile-corrosion tests. It is intended that the outcome of this research will be the establishment of a systematic multi-scale multi-physics experimental and numerical protocol for predicting aged structural resilience. INTRODUCTION Since the 1950s the construction time of ships and oil rigs has been significantly reduced due to advances in welding and maintenance technologies, whereas the operating lifespan of assets have doubled or even tripled. Yet marine platforms operate within one of the harshest environments for mechanical structures. Although various prevention systems are currently applied, corrosion damage, often in a localized area, is still occurring due to protective coating breakdown, insufficient cathodic protection, inappropriate material selection and lack of prompt maintenance. For example, the so-called “super rust” found in oil tankers has a corrosion rate 5 to 7 times higher than the normally accepted uniform corrosion rate [1]. Moreover, severe corrosion damage is frequently found along weldments and heat-affected zones, which facilitate cracking and deteriorate the structural integrity [2, 3]. Lack of a firm understanding of the operational environment directly increases the unpredictability of corrosion, hence leading to time consuming and costly correcting measures when either the infrastructure or vessel is operational [4]. In 2011, the global cost of marine corrosion was estimated to be a staggering $7.5 billion for new constructions and $5.4 billion for repair and maintenance [5]. This leads to an annual loss of 60 million tons of steel of which 25–30% could be saved if corrosion prevention strategies are appropriately employed. Additionally, the introduction of more flexible surveying schedules could help to reduce maintenance costs and increase operating times. Through maritime platforms’ service life, predicting the material degradation process is inherently a complex understating for that even a single type of marine corrosion often involves multiple stages, each of which has different elementary steps or processes across several length scales. Moreover, the degradation process is at the often multi-physics/multi-scale and involves geometric complexities for instance physical joints and crevices, as well as mechanical factors such as in-service loading and weld-induced residual stresses that affect corrosion initiation and kinetics. The nature of protective films and corrosion product scales can also be affected by excessive structural deformation. This leads to significant challenges when establishing reliable structural performance and inspection/repair strategies. The key research disciplines include corrosion prediction and structural analysis. The majority of the existing corrosion prediction models used in engineering applications are in statistical/empirical forms based on field measurements. Although such models have been published in recent years [6, 7], the most up-to-date corrosion data sets that are available in the open literature were dated back in 2005 and 2008 for singlehull oil tankers and bulk carriers respectively, with only 1404