University of Illinois at Chicago
Heidary_Zahra.pdf (4.8 MB)
Download file

Numerical Modeling of Wave Propagation at Large Scale Damaged Structures For Quantitative AE

Download (4.8 MB)
posted on 2015-02-27, 00:00 authored by Zahra Heidary
The Acoustic Emission (AE) method is a nondestructive testing method that relies on the waves emitted from the localized permanent deformation. The method can detect, locate and characterize the damage in large-scale structures. Unfortunately, the AE method is not being used to its full advantage because of a problem in detecting the induced micro-damage signals, which can be obscured by mechanical noise and the complexity of the measurement leading to challenges of repeatability and extracting quantitative features (e.g., damage mode, size, direction). The goal of this research is to understand the quantitative significance of the AE signatures through effective finite element models combining the components of source, structure and sensor into the model. The domain is modeled using spectral element, which reduces the required degrees of freedom significantly; the boundaries are modeled with Perfectly Matched Layer (PML) to absorb the waves, similar to large-scale structures. The numerical models are validated with experimental measurements. The numerical model together with novel mathematical formulation of axisymmetric pipe under non-axisymmetric loading (e.g., leak) allowed understanding wave propagation in long-range pipelines in order to determine reliable sensor spacing. The transfer function of the AE sensor (typically piezoelectric) is coupled with the solid model to obtain the electrical displacement of the sensor under given excitation and understand the influence of the sensor response to the output signal.



Ozevin, Didem


Civil and Materials Engineering

Degree Grantor

University of Illinois at Chicago

Degree Level

  • Doctoral

Committee Member

Foster, Craig D. Indacochea, J. Ernesto Shabana, Ahmed McNallan, Michael J.

Submitted date



  • en

Issue date


Usage metrics


    No categories selected