NDE Reliability using Laboratory Induced Natural Fatigue Cracks

Successful implementation of damage tolerance methodology requires estimation of the reliability of the currently practiced non-destructive testing ( NDT) techniques in the aero-engine industry. Reliability of the NDT techniques is measured in terms of probability of detection (POD). POD is dependent on m aterial, geometry along with size, shape, location and type of the defect. Experime ntal estimation of POD usually requires large number of service expired aero-engine turbine disks. Alternatively, the POD studies can also be carried out usin g laboratory induced samples (as per MIL-HDBK-1823A) containing various sizes o f defects. In the current study, 55 mm x 10 mm x 10 mm nickel based super alloy s mples were extracted from a mid-life service expired turbine disc. A 2 mm radius circular notch was created at the centre of the specimen so as to represent the true stress concentration factor present near the bolt hole location of a turbine disc. Sa mples were subjected to repeated stress fatigue cycling using a 3-point bend set up with a constant load of -6.5 kN and stress ratio of R = 0.1. All these sam ples were subjected to constant number of fatigue cycles (1.05 x 10 ) under room temperature high cycle fatigue (HCF) conditions. 50 different cracks ranging from 0.1 mm – 2.0 mm sizes were observed in these fatigued samples. Fluorescent penetrant and eddy current inspection tests were carried out on the specimens for qualitative (Hi /Miss) defect indications. Further, Hit/Miss data obtained from both the techniques was analysed and the POD curve was plotted. Finally, the a 90/95 (flaw detection with 90 % probability and 95 % confidence) values of penetrant and eddy curren t inspection studies of naturally grown fatigue cracks were reported to be 1.13 mm and 0.98 mm, respectively.