Ph.D. Thesis Defense

By

Olivier Mesnil

(Advisor: Prof. Massimo Ruzzene)

Sparse reconstruction and analysis of guided wavefields for damage detection and quantification

1:00 PM, Thursday, May 19th, 2016

Montgomery Knight Building

Room 317

            This thesis presents the study of novel guided wave-based techniques that locate, quantify and analyze defects in composite materials and metals. These techniques find themselves at the intersection of Structural Health Monitoring (SHM) and Non Destructive Evaluation (NDE) due to their use of both guided wave methods and Guided Wavefield Imaging (GWI) techniques. The primary goals of NDE are limited to the local detection and the evaluation of damages. NDE is often characterized by off-line inspection requiring a cumbersome equipment. On the other hand, SHM generally aims at monitoring the condition of the structure during its regular operating cycle, ideally with embedded sensors. Guided wave methods are commonly considered to be part of SHM methodologies and are widely used for the detection of flaws in plate-like structures. Guided Wavefield Imaging is a technique relying on the detection of images corresponding to the time evolution guided waves propagation in the structure. Due to the high sensibility of guided waves to internal defects and the amount of information they include, wavefields have the potential to provide extensive information regarding the structural component under consideration. However the main limit to GWI if the time consuming guided wavefield acquisition process.

            The objectives of this dissertation is to develop novel GWI techniques able to detect, locate and quantify defects in metals and composite materials while reducing the acquisition time. In addition, the work attempts to link guided wave signatures to damage parameters in order to provide strength estimates. Two complimentary techniques are developed to achieve the objectives. The first technique, namely the Sparse Wavefield Reconstruction (SWR) reduces the acquisition time required to detect a wavefield. The approach employs sparse measurements, i.e. fewer measurements than required by the traditional sampling criteria, scattered within the region of inspection. The second technique, namely the Frequency Domain Instantaneous Wavenumber (FDIW), focuses on the quantification of a defect through local measurement of the wavenumber. The combination of these two techniques provides an inspection methodology that reduces the inspection time and provides detailed defect quantification. Finally, a novel methodology bridging guided wave measurements and residual life prediction is presented. This study measures the constitutive damage parameters of composite panels in order to obtain residual properties estimates through damage mechanics numerical models. This methodology is used to measure the constitutive damage variable of a numerically simulated damage and show promising results for an experimental application.

Advisory Committee:

Prof. Massimo Ruzzene (AE, Advisor)

Prof. Krish Ahuja (AE)

Prof. George Kardomateas (AE)

Prof. Kamran Paynabar (ISyE)

Prof. Jennifer Michaels (ECE)