Author's School

Arts & Sciences

Author's Department

Mathematics

Document Type

Conference Proceeding

Publication Date

2-10-2016

Originally Published In

AIP Conf. Proc. 1706, 120002 (2016); http://dx.doi.org/10.1063/1.4940587 Conference date: 26–31 July 2015

Abstract

This study compares different approaches for imaging a near-surface resin-rich defect in a thin graphite/epoxy plate using backscattered ultrasound. The specimen was created by cutting a circular hole in the second ply; this region filled with excess resin from the graphite/epoxy sheets during the curing process. Backscat-tered waveforms were acquired using a 4 in. focal length, 5MHz center frequency broadband transducer, scanned on a 100 × 100 grid of points that were 0.03 × 0.03 in. apart. The specimen was scanned with the defect side closest to the transducer. Consequently, the reflection from the resin-rich region cannot be gated from the large front-wall echo. At each point in the grid 256 waveforms were averaged together and subsequently used to produce peak-to-peak, Signal Energy (sum of squared digitized waveform values), as well as entropy images of two different types (a Renyi entropy, and a joint entropy). As the figure shows, all of the entropy images exhibit better border delineation and defect contrast than the either the 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. The first is a reflection of the insonifying pulse from a stainless steel reflector. The second is an approximate optimum obtained from an iterative parametric search. The joint entropy images produced using this reference exhibit three times the contrast obtained in previous studies.

Comments

© 2016 AIP Publishing LLC but noted as open access content

AIP Conf. Proc. 1706, 120002 (2016); http://dx.doi.org/10.1063/1.4940587
  • Conference date: 26–31 July 2015
  • Location: Minneapolis, Minnesota

DOI

10.1063/1.4940587

Share

COinS