Most demanding components – for example heavy rotor forgings for the energy sector – are inspected with volumetric and surface detection NDE methods – either manually or using automated systems. The data collected during automated inspection gets processed and analyzed. Similarly, the inspector analyzes the information he sees during manual inspections. The results, like size and position of indications, and the inspection conditions, like sensitivity information and sound attenuation, are finally condensed into reports both for manual and automated inspections. Finally, those reports get printed, signed and archived. However, the data reported in most of those reports is a treasure which needs to be exploited by the NDE community. One key application for such a smart data analysis is probabilistic fracture mechanics.
We suggest a better utilization of the NDE inspection data by the implementation of a state of the art data-base for easy access and latest statistical analysis technologies. We report on first results of a newly developed smart data analysis approach on the example of its use for probabilistic fracture mechanics for heavy duty rotor forgings, one of the most demanding components from an inspection and design standpoint. Preliminary analysis, including probability of detection, indication distribution, as well as detection sensitivity comparisons are presented.