High sensitivity imaging of resin-rich regions in graphite/epoxy laminates using joint entropy

Abstract

The continuing difficulty of detecting critical flaws in advanced materials requires novel approaches that enhance sensitivity to defects that might impact performance. This study compares different approaches for imaging a near-surface resin-rich defect in a thin graphite/epoxy plate using backscattered ultrasound. The specimen, having a resin-rich void immediately below the top surface ply, was scanned with a 1 in. dia., 5 MHz center frequency, and 4 in. focal length transducer. A computer controlled apparatus comprised of an x-y-z motion controller, a digitizer (LeCroy 9400A), and an ultrasonic pulser/receiver (Panametrics 5800) was used to acquire data on a 100 ×100 grid of points covering a 3 × 3 in. square. At each grid point 256 512-word, 8-bit backscattered waveforms, were digitized, signal averaged, and then stored on computer for off-line analysis. The same backscattered waveforms were used to produce peak-to-peak, signal energy, as well as entropy images. All of the entropy images exhibit better border delineation and defect contrast than the either peak-to-peak or signal energy. The best results are obtained using the joint entropy of the backscattered waveforms with a reference function. Two different references are examined: a reflection from a stainless steel reflector, and an approximate optimum obtained from an iterative parametric search. The joint entropy images produced using the optimum reference exhibit ~3 times the contrast obtained in previous studies.