Date of Award

2019

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

First Advisor

Dr. Mannur Sundaresan

Abstract

Acoustic emission signal in aluminum plates and composite specimens are examined in this dissertation. The first part of the dissertation examines the symmetric lamb wave modes generated by the spontaneous appearance of a central microcrack in an aluminum plate. The detected acoustic emission amplitude of these waves with radial and angular position of the sensing location was analyzed using finite element analysis. An important aspect of this study is the detailed characterization of the shear horizontal mode of the Lamb wave that is generated by simulated crack growth. This analysis indicates that the amplitude of the shear horizontal mode remains larger than the fundamental symmetric mode generated by a simulated through-the-thickness crack at the center of the plate. Design of new sensors capable of efficiently isolating and detecting shear horizontal mode was examined. The final sensor configuration chosen was reasonably successful in the experiments that employed simulated acoustic emission sources launched into an aluminum plate using pulsing transducer excited by a triangular pulse. However, this sensor was not able provide clean signals for detecting crack growth related shear horizontal mode during a fatigue crack growth experiment. The second part of the dissertation was on the development of structural health monitoring technique capable of assessing damage in composite structural elements. Acoustic emission signals generated during quasi-static loading of cross-ply tensile specimens were examined to relate each category of signal to the type of failure event and the failure mode. Based on the frequency content of the signal these signals were categorized. As reported by a number of researchers in the past, the critical damage that occurs before the catastrophic failure are the formation of clusters of multiple fiber breaks. Currently, such clusters of fiber breaks are identified after the specimen is unloaded, using scanning electron microscopy, or through 2 miniature tensile specimens loaded inside the chamber of a computed tomography instrument. In this study it is hypothesized that nearly identical acoustic emission waveforms are generated by clusters of fibers failing together near the final failure. Hence such signals may be useful for forewarning the impending failure. This hypothesis was examined for both undamaged specimens as well as specimens that were subjected to transverse impact. In both cases, rapid increase in the clusters of identical acoustic emission waveforms were seen and were successful in predicting impending failure. This technique needs to be validated through more extensive tests so that it can be scaled up to monitor real life structures such as primary structures of aerospace vehicles.

Download

Share

COinS