Through Life Reliability of a Bulk Carrier

Tsarouchas, Ioannis (2001) Through Life Reliability of a Bulk Carrier. PhD thesis, University of Glasgow.

Abstract

A methodology for a through life structural assessment of a ship is proposed and applied to a bulk carrier. The objective of the analysis is to calculate the probability of failure of the corroding vessel, having many initial defects which grow through fatigue. The assessment requires bringing together many different analyses. Loading is a key input in the analysis. There are two important loads that need to be considered: 1) Fatigue loading, which causes cracks to grow. A fatigue analysis provides useful information on fatigue loading (stress ranges) on many different structural details found in a bulk carrier. The operational profile of the vessel is specified together with the loading conditions and the service speed. Modeling the wave environment is done using a sea scatter diagram. From the fatigue analysis a selection of details to consider for the analysis is possible (i.e. details with low fatigue lives). 2) Extreme loading in conjunction with cracks causes fracture. Short-term statistical analysis on wave environment is performed to estimate extreme loads. Other loads are the still water loading and self-equilibrating stresses of which welding residual stresses have been included. Fracture mechanics are used to perform the crack extension calculations. A simplified model based on Paris' law is used, modified to include threshold and corrosion effects. Suitable crack growth constants are selected that simulate crack extension in a corrosive environment with a reasonable accuracy. Another fundamental part of the analysis is to establish a fracture criterion. A suitable criterion is based on a failure assessment diagram described in PD 6493. This accounts for the plastic zone which develops in front of the crack tip, in addition to the simplified fracture criterion based on linear elastic fracture mechanics. By using this method, the need to use non-linear analysis (J-Integral estimation) is avoided. A bulk carrier has many possible structural details where initial welding defects can develop into cracks and propagate. Each crack will grow at a different rate. Crack growth parameters are used taken from published references. Information on the stress intensity factors, for typical cracked details is obtained either from published references or by the use of FEA. Most parameters that are modelled as random variables are specified and their uncertainty is described by a probability distribution defined by a mean value and a coefficient of variation and type of distribution. Reliability analysis is performed using the Monte Carlo simulation method. This was selected because it is the most flexible method to perform complicated time- dependent reliability analysis (because of the crack propagation) and combine all necessary parts into one computer program. Because of the large number of locations to consider, (e.g. for a realistic case of a ship structure there are thousands of cracks present, and each crack must be treated separately in the simulation), there are limitations on the use of the simulation method (it requires a lot of computer memory and computational time). Because of these restrictions, the simulation is limited to approximately 100 cracks. A real ship structure requires many thousands of cracks to be considered so a simplified methodology is established which uses results (load and strength distributions) from the simulation method (with a limited number of cracks), and is capable of evaluating the failure probability of the whole ship. For the purpose of demonstrating the methodology, simplified assumptions are made for the input parameters. Cracks are distributed over the entire vessel, subjected to different loading, having different material properties (for crack growth calculations). Loading variation along the structure is based on simplified assumptions. Results of the method are compared with the simulation (for a limited number of cracks) and with some actual statistical data available from various sources e.g. classification societies or other organizations related to ships. Although the absolute values from the analysis may not be accurate (since simplified assumptions for the input data have been made), however the trend of the results are of interest and in reality the results are the best estimate available. By performing several analyses it is possible to study the effects of corrosion, inspection, sea route and loading conditions on the reliability of the vessel. At the design stage requirements and methods for inspection can be balanced against the provision of additional steel, better details, better steel toughness grades. When more realistic data is available such a methodology can be used to schedule optimum inspection intervals, by maintaining the reliability levels to a specified target value. This will result in a more efficient, safe and economic way of vessel operation. It is important for the application of the methodology to have a database of input parameters that can be used by consultants, ship owners, or classification societies when undertaking this type of analysis. This includes data on crack growth parameters, fracture toughness, corrosion rates, initial crack sizes and distribution of cracks. Assumptions made on the physical models can also be improved, such as loading estimation (fatigue and extreme), crack propagation process, accuracy of the reliability method to refine the simplified method.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Adviser: P K Das
• Keywords: Naval engineering
• Date of Award: 2001
• Depositing User: Enlighten Team
• Unique ID: glathesis:2001-75747
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 19 Dec 2019 09:15
• Last Modified: 19 Dec 2019 09:15
• URI: https://theses.gla.ac.uk/id/eprint/75747