A multivariate model to assess the probability of detection and sizing of defects in aluminum panels using eddy current inspections. (December 2018)

• Record Type: Journal Article
• Authors: Barrett, Adam; Smith, Reuel; Modarres, Mohammad
• Is Part Of: Engineering failure analysis. Volume 94(2018)
• Page Start: 182
• Page End: 194
• Publication Date: 2018-12
• Subjects: System failures (Engineering) -- Periodicals; Fracture mechanics -- Periodicals; Reliability (Engineering) -- Periodicals; Pannes -- Périodiques; Rupture, Mécanique de la -- Périodiques; Fiabilité -- Périodiques; Fracture mechanics; Reliability (Engineering); System failures (Engineering); Periodicals; Electronic journals; 620.112
• DOI: 10.1016/j.engfailanal.2018.07.028
• Languages: English
• ISSNs: 1350-6307

Abstract

Abstract: Typical inspection capability models are two-parameter models that estimate the probability of detection of certain size defects, flaws or cracks under controlled inspection conditions. This paper expands traditional detection models, including customary probability of detection and sizing models, to a multivariate model that includes additional factors that affect inspection outcome. To better assess inspection system capabilities, two procedures are examined. The first procedure involves the traditional model of detection and sizing based on a set of inspection system data which includes hit-or-miss data, sizing parameters, and various conditions pertaining to the inspection process. The second procedure implements a multivariate model of the flaw detection and sizing. The probability of detection, sizing, and certain shaping factors based on various inspection factors and conditions are introduced and modeled to demonstrate the impact of inspection system capabilities and conditions in the field. Highlights: An inspection system study comparing two approaches that examine how different input factors affect flaw detection modeling An approach that considers the effectiveness of detection probability, flaw sizing models, and model error An approach that applies Bayesian estimation, a joint flaw sizing/detection model, Gaussian process, and particle filtering A sensitivity analysis was implemented examining several shaping factors that affect the flaw size detection capabilities of the inspectors under study. Some shaping factors affected the apparent flaw size more than others.