Terahertz non-destructive sensing of layered materials with curved surfaces

Terahertz (THz) imaging has shown great potential for nondestructive evaluation (NDE) of a wide variety of manufactured products. Its ability to penetrate many non-polar dielectrics allows tomographic imaging of an object's 3D structure. The authors' previous work demonstrated the ability to produce 3D THz tomographic images of objects using a single data set collected in a 2D synthetic aperture configuration. This ability was recently expanded to allow imaging to within objects with curved boundaries by employing a split-step Fourier method to numerically back propagate the measured fields to the set of image planes that comprise the 3D tomographic image. This work presents new experimental results that demonstrate the split-step Fourier method's ability to correct for refraction and accurately quantify the thickness of a curved dielectric layer.