Validation of the Rotating UT Probe for In-Service Inspections of Freight Solid Axles by Means of the MAPOD Approach

The present scenario of freight applications in Europe is characterized by a trend towards rolling stock renewal and a new approach to periodical inspection and maintenance; this is a consequence of some evidences coming from service and showing potential inconsistency in general safety levels. A specific NDT technique, based on more than 40 years of service experience on passenger wheel-sets inspection, is here proposed for periodical NDT on freight wheel-sets. The ultrasonic testing technique is based on a rotating probe-holder to be applied at the ends of the mounted axle and able to inspect its critical sections with different angled probes. To inspect such critical sections such as press fits and geometrical transitions, the testing device is characterized by a set of ultrasonic transducers emitting longitudinal waves at a typical nominal frequency of 4 MHz and at different angles into the axle. The structural integrity of the wheel-set is then based, together with other factors, on the determination of the Probability of Detection (POD) curve of the rotating probe. An experimental campaign to derive the POD curve can be costly and time consuming, especially considering the determination of its confidence range. For this reason, in the present research, a Model-Assisted POD (MAPOD) approach was considered. Initially, different sets of artificial defects, obtained by EDM, were introduced on the external surface of three solid axles made of A1N steel grade. All the defects were then inspected by a rotating probe, with the aim to derive their response, considering 4 MHz and 2.25 MHz transducers. The effect of the different time of flight characterizing the defects located at the different sections of the axles was, instead, determined by suitable numerical simulations carried out using the CIVAnde dedicated software package. Combining experimental and numerical results allowed to derive the MAPOD curve for the considered rotating probe with less effort with respect to the traditional experimental approach and to optimize the set-up of the rotating probe itself. As a result, the MAPOD curves of the rotating probe device are compared to the POD of standard UT inspection from axle end and a simulation of the reliability of inspection intervals is presented.