Bayesian analysis of flaw sizing data of the NESC III exercise

Luca Gandossi (Corresponding Author), Kaisa Simola

Research output: Contribution to journalArticleScientificpeer-review

4 Citations (Scopus)

Abstract

Non-destructive inspections are performed to give confidence of the non-existence of flaws exceeding a certain safe limit in the inspected structural component. The principal uncertainties related to these inspections are the probability of not detecting an existing flaw larger than a given size, the probability of a false call, and the uncertainty related to the sizing of a flaw. Inspection reliability models aim to account for these uncertainties. This paper presents the analysis of sizing uncertainty of flaws for the results of the NESC III Round Robin Trials on defect-containing dissimilar metal welds. Model parameters are first estimated to characterize the sizing capabilities of various teams. A Bayesian updating of the flaw depth distribution is then demonstrated by combining information from measurement results and sizing performance.
Original languageEnglish
Pages (from-to)654-662
JournalInternational Journal of Pressure Vessels and Piping
Volume83
Issue number9
DOIs
Publication statusPublished - 2006
MoE publication typeA1 Journal article-refereed

Fingerprint

Defects
Inspection
Dissimilar metals
Welds
Uncertainty

Keywords

  • Inspection reliability
  • NDT
  • Bayesian modelling
  • Probability models for flaw sizing
  • NESC

Cite this

Gandossi, Luca ; Simola, Kaisa. / Bayesian analysis of flaw sizing data of the NESC III exercise. In: International Journal of Pressure Vessels and Piping. 2006 ; Vol. 83, No. 9. pp. 654-662.
@article{b2f6342e30994d9b9f1a4ebf98de957e,
title = "Bayesian analysis of flaw sizing data of the NESC III exercise",
abstract = "Non-destructive inspections are performed to give confidence of the non-existence of flaws exceeding a certain safe limit in the inspected structural component. The principal uncertainties related to these inspections are the probability of not detecting an existing flaw larger than a given size, the probability of a false call, and the uncertainty related to the sizing of a flaw. Inspection reliability models aim to account for these uncertainties. This paper presents the analysis of sizing uncertainty of flaws for the results of the NESC III Round Robin Trials on defect-containing dissimilar metal welds. Model parameters are first estimated to characterize the sizing capabilities of various teams. A Bayesian updating of the flaw depth distribution is then demonstrated by combining information from measurement results and sizing performance.",
keywords = "Inspection reliability, NDT, Bayesian modelling, Probability models for flaw sizing, NESC",
author = "Luca Gandossi and Kaisa Simola",
year = "2006",
doi = "10.1016/j.ijpvp.2006.06.002",
language = "English",
volume = "83",
pages = "654--662",
journal = "International Journal of Pressure Vessels and Piping",
issn = "0308-0161",
publisher = "Elsevier",
number = "9",

}

Bayesian analysis of flaw sizing data of the NESC III exercise. / Gandossi, Luca (Corresponding Author); Simola, Kaisa.

In: International Journal of Pressure Vessels and Piping, Vol. 83, No. 9, 2006, p. 654-662.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Bayesian analysis of flaw sizing data of the NESC III exercise

AU - Gandossi, Luca

AU - Simola, Kaisa

PY - 2006

Y1 - 2006

N2 - Non-destructive inspections are performed to give confidence of the non-existence of flaws exceeding a certain safe limit in the inspected structural component. The principal uncertainties related to these inspections are the probability of not detecting an existing flaw larger than a given size, the probability of a false call, and the uncertainty related to the sizing of a flaw. Inspection reliability models aim to account for these uncertainties. This paper presents the analysis of sizing uncertainty of flaws for the results of the NESC III Round Robin Trials on defect-containing dissimilar metal welds. Model parameters are first estimated to characterize the sizing capabilities of various teams. A Bayesian updating of the flaw depth distribution is then demonstrated by combining information from measurement results and sizing performance.

AB - Non-destructive inspections are performed to give confidence of the non-existence of flaws exceeding a certain safe limit in the inspected structural component. The principal uncertainties related to these inspections are the probability of not detecting an existing flaw larger than a given size, the probability of a false call, and the uncertainty related to the sizing of a flaw. Inspection reliability models aim to account for these uncertainties. This paper presents the analysis of sizing uncertainty of flaws for the results of the NESC III Round Robin Trials on defect-containing dissimilar metal welds. Model parameters are first estimated to characterize the sizing capabilities of various teams. A Bayesian updating of the flaw depth distribution is then demonstrated by combining information from measurement results and sizing performance.

KW - Inspection reliability

KW - NDT

KW - Bayesian modelling

KW - Probability models for flaw sizing

KW - NESC

U2 - 10.1016/j.ijpvp.2006.06.002

DO - 10.1016/j.ijpvp.2006.06.002

M3 - Article

VL - 83

SP - 654

EP - 662

JO - International Journal of Pressure Vessels and Piping

JF - International Journal of Pressure Vessels and Piping

SN - 0308-0161

IS - 9

ER -