POD and GUM Universal Methods for Making Safety Measurable

Abstract

Various non-destructive testing (NDT) methods such as ultrasounds (UT) or eddy current (ET) have been established for in-service inspections (ISI) or for condition assessment in different kind of industries such as in the nuclear or aerospace business. Another example for a common NDT inspection task is the detection of tendon ducts in the field of civil engineering (CE) using Ground Penetrating Radar (GPR) as well as the determination of its lateral and depth position. Therefore, the detection limits of the used inspection system, which depends on the depth position and the number of tendon ducts and distances between them, has to be well known. One approach to determine the limits of detection is the application of the POD method (probability of detection) as a universal procedure, which excludes the human factor and calculates the a_90/95 value and other characteristic parameters. With this information, different inspection systems can be compared objectively to choose the most suitable NDT-system for each individual inspection task. To distinguish between a real variation in the tendon duct position and the accuracy of the inspection system, the knowledge about the uncertainty of measurement is required. To determine the accuracy of the selected NDT-system, the GUM procedure (guide to the expression of uncertainty in measurement) has been established and provides a statistical evaluated result in form of the measurement result and its expanded uncertainty. This article introduces a procedure using the example of tendon duct detection (POD) and depth position description (GUM) in concrete with Ground Penetrating Radar (GPR). Finally, the universal application of both methods (POD and GUM) in different fields of industries is illustrated by some examples.