A Bayesian quantitative nondestructive evaluation (QNDE) approach to estimating remaining life of aging pressure vessels and piping

Abstract

In this paper, we address a generic problem of estimating the remaining useful life of an aging structure by a Bayesian quantitative nondestructive evaluation (Q-NDE) approach, where the key emphasis is in first obtaining as much as possible the relevant information on the current material properties of the aging structure, and then applying the NDE-based information as input to a stochastic multiple-crack-growth fatigue life model with judgment-based a priori assumption of crack length distributions (Bayesian). A numerical example using synthetic NDE data for a typical class of high strength steel pressure vessels or piping is presented to illustrate the novel approach.

A Study of Quantifying Thickness of Ferromagnetic Pipes Based on Remote Field Eddy Current Testing

“…In the Eddy Current Testing (ECT) of ferromagnetic pipes, including RFECT of ferromagnetic pipes, achieving wall thickness quantification is always the first priority [ 11 , 12 , 13 , 14 ]. Accompanied by the progress of computer and computing technology, advanced computing methods (for example, Bayesian network [ 15 ], artificial neural networks [ 16 ] and Support Vector Machine (SVM) [ 17 ]) are employed to reconstruct and quantify defects. The authors also studied and compared evaluation methods of pipe defects based on nonlinear fitting, neural networks and SVM.…”