Terahertz Imaging System for Automatic NDE and Detection of Foreign Objects in Composite Components

This proposal describes an approach for the use of terahertz band radiation, or T-rays, for nondestructive evaluation (NDE) and detection of foreign objects in manually constructed composites. Contaminates introduced during the fabrication phase of composite airframe parts contribute to high rejection rates. A method of identifying foreign objects embedded in a lay-up before curing of the part, and allowing their removal, would reduce the financial penalties associated with manufacturing composites. T-rays have many advantages over traditional NDE techniques, including being non-ionizing radiation and the ability to "see-through" many opaque materials. Due to the unique approach to constructing terahertz images, new processing techniques are applied in order to identify and extract information from the images relevant to detecting foreign objects. A multi-resolution decomposition is coupled with Bayesian classifier to learn the characteristics of foreign objects embedded within various composite types and allow automatic detection.