Model Assisted Classification for Probability of Detection (MACPOD) using Sonic Infrared Inspection

The experimental POD procedure requires a large amount of empirical data consisting of repeated inspections of the same standard specimens by various NDE operators under various environmental and operational conditions. Sonic Infrared (IR) inspection has been successfully demonstrated and implemented at the Air Force depot to inspect turbine engine blades for serviceability. To reduce the requirement for costly POD assessments and help extend the usage of POD of the Sonic IR technology, Intelligent Automation, Inc. (IAI) with its partner proposes to develop and test a novel simulation based POD grouping tool called MACPOD (Model Assisted Classification for Probability of Detection) for Sonic IR inspection of turbine engine compressor blades. The proposed tool will develop a model-assisted POD methodology for assessing engine blades reliability with Sonic IR technology. It will derive a physics-based numerical or machine learning model to classify the groups of turbine engine blades where existing POD data can be extended for their inspection with Sonic IR. The resultant MACPOD tool will utilize simulation to classify the POD analysis for turbine engine blades to reduce the standard Sonic IR POD tests under different excitation parameters.