A Stochastic Model For Determining Inspection Intervals For Large Marine Vessels

摘要

This paper presents a stochastic model that accounts for failure probability as a function of inspection frequency and effectiveness. A total ownership cost (TOC) model is presented. The model is then applied to a cargo vessel. Stress-strength relationships predict the failure likelihood of details or critical points within the structure. These individual points are combinatorially combined into a single higher level model. An inspection plan together with the effectiveness of inspection with and without NDE is then used to determine the overall failure probability as a function of time and inspection intervals. Inputs into the model include the effectiveness of the NDE, the strength degradation and the stress distribution as a function of time, the cost of the NDE method, the effectiveness of the NDE method, the effectiveness of the inspection without NDE, and the cost of failure. Cost benefits can then be determined on the basis of a decreased failure likelihood with better detection methods or, alternatively, increased inspection intervals given a constant acceptable failure probability. Presentation of a prototype of this methodology implemented in MATLAB and using a military cargo ship as an example is presented.