Volume 16, Issue 3 (2016)                   MCEJ 2016, 16(3): 203-215 | Back to browse issues page

XML Persian Abstract Print


Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Moarefzadeh M R. Reliability of Marine Steel Structures Against Corrosion. MCEJ 2016; 16 (3) :203-215
URL: http://mcej.modares.ac.ir/article-16-3397-en.html
Assistant Professor/The University of Imam Hossein
Abstract:   (4060 Views)
Abstract Reliability analysis of steel structures subject to seawater corrosion is of considerable interest for coastal and offshore marine steel structures. These are often very expensive and have high consequential costs and implications should failure occur. Since corrosion of steel structures causes deterioration of structural strength, usually gradually with time, safety assessment is of considerable importance for new structures (those in the design stage) and also for those which are already in operation. Marine corrosion is a complex phenomenon and subject to various influencing factors each of which has its own inherent uncertainty. In any safety assessment, in principle, the uncertainty of each factor should be studied and taken into account. Since such an action is too difficult, in practice some test programs are normally conducted and all uncertainties caused by different factors are assumed to be included in the relevant corrosion measurements. In addition, in any corrosion reliability analysis for steel structures exposed to seawater, two different models must be taken into consideration: (1) A physical model indicating general corrosion behaviour as a function of exposure time and (2) A stochastic model describing probabilistic treatment of uncertainties observed in real corrosion data. The first has been traditionally treated by invoking a simple power law and in particular a linear relationship. However, using realistic long-term data, validity of such a model has recently been challenged. The second model (i.e. probabilistic modelling of corrosion process) has been dealt with in literature in different approaches, including either taking the corrosion annual rate as a random variable or proposing a stochastic process such as Gamma process. This is usually proposed as a general structural deterioration process. The second approach provides, doubtlessly, better treatment of corrosion uncertainties; however it can be shown unfortunately that the Gamma process is unable to reflect the corrosion uncertainties in some circumstances. In this paper, two sets of corrosion data collected in different seawaters around the world with different temperatures are used. This requires processing of data in such a way that the data sets remain consistent with each other and that outcome is data that can be considered as belonging to one statistical population. Herein, first, a simple algorithm is proposed to transform the whole data to one common temperature. Second, a novel Markov-chain based model is developed which meets long term second-order corrosion statistics (i.e. means and standard deviations of corrosion losses). It is based on a corrosion model that previously has been calibrated extensively to field observations of corrosion and to literature-reported realistic data. Although actual long-term field observations of marine corrosion of steel are scarce, it is shown that particularly for the standard deviation the new model is well capable to be consistent with the long-term data. It is noted that herein, only the corrosion data collected in marine immersion zones are considered (i.e. those taken in splash zones and atmospheric zones are not considered). Further, only general corrosion (i.e. not pitting corrosion) is accounted for herein. These issues, obviously, have to be addressed separately.
Full-Text [PDF 842 kb]   (3243 Downloads)    
Article Type: Research Paper | Subject: -------
Received: 2014/04/14 | Accepted: 2015/06/17 | Published: 2016/07/22

Add your comments about this article : Your username or Email:
CAPTCHA

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.