Improved Methodology for Inspection Reliability Assessment for Detecting Synthetic Hard Alpha Inclusions in Titanium

This paper presents a progress report on the development of model-based methodology to estimate Nondestructive Evaluation (NDE) capability. Previous work on this project was reported in Meeker et al. [3, 4]. The methodology uses combinations of physical modeling of an inspection process, along with laboratory and production data, to estimate Nondestructive Evaluation (NDE) capability. The methodology is based on a physical/statistical prediction model and will be used to predict NDE capability in terms of Probability of Detection (POD), Probability of False Alarm (PFA), Probability of Indication (POI) and Receiver Operating Characteristic (ROC) curves. The physical model explains and allow predictions for the effects of making changes to the inspection setup (e.g. transducer properties and scan increment). The statistical/empirical model will quantify unexplained variability, adjust for potential model bias, and provide a means for obtaining corresponding uncertainty intervals. The work has been motivated by the need for methods to predict ultrasonic (UT) inspection POD for detecting hard-alpha and other subsurface flaws in titanium using gated peak detection. This is a challenging problem because the inspection must detect very complex subsurface flaws in the presence of significant “material” noise. The underlying framework of the methodology should, however, be general enough to apply to other NDE methods. This paper describes recent work based on application of the new methodology to the detection of synthetic hard alpha flaws in titanium alloys. In particular we describe and illustrate methods to assess the effect that changes in scan plans and gate width will have on POD