Search

Early demonstrations of the use of models in the U.S. and U.K. to assist in the determination of NDE reliability and probability of detection (POD), as dictated by the advent of damage tolerant design, will be reviewed. A major motivation in the U.S. was the adoption of POD as the primary metric for effectiveness of an inspection and the empirical observation that field and depot inspections were not adequate so support the desired structural integrity. In parallel, a series of round robin trials in Europe raised issues with the effectiveness of nuclear power plant inspections, and the resulting concerns became an important part of the 'Public Inquiry' in the UK regarding the construction of the Sizewell B power plant. In each country, physics-based models of the inspection process became an important part of the response. This paper describes these early developments in the use of models to assess inspection reliability, contrasts the philosophies and approaches taken, and notes how these provide the origins of NDE simulation activities today. [ABSTRACT FROM AUTHOR]

In the nineteen-sixties, a significant limitation of NDI and NDE became apparent with the advent of fracture mechanics. Fracture mechanics requires quantitative information on defects, which has to be obtained from quantitative non-destructive testing. A DARPA Program directed by Don Thompson provided the point of departure for the journey to put the Q with NDE. The DARPA Program, and subsequent DOD, FAA and industrial programs produced seminal results for diagnostics and prognostics. In diagnostics, measurement models, probability of detection considerations and techniques of defect characterization were developed, which were complemented by damage evolution laws, probabilistic failure analysis and damage progression estimates, for methods of prognostication. The new results in QNDE naturally led to the concept of structural health monitoring (SHM), whereby sensors are permanently installed on structures. An SHM system can provide on-demand (or continuous) information on the state of a structure, so that an assessment of the structural integrity can be made at any time, and timely remedial actions can be taken. In this paper, we review the development of QNDE towards SHM. Sensor development, data processing, materials engineering and solid mechanics play dominant roles in both the diagnostic and the prognostic components of SHM. A probabilistic approach is essential, as will be shown by examples of pre-crack fatigue damage, crack growth and optimization of an inspection schedule. [ABSTRACT FROM AUTHOR]