Date of Award

Spring 2020

Document Type

Open Access Dissertation


Mechanical Engineering

First Advisor

Victor Giurgiutiu


This dissertation is organized into major parts. In Part I of the dissertation, plate guided waves due to an acoustic emission (AE) event are analytically studied in the context of seismic moment tensor sources. The guided wave propagation elastodynamic equation corresponding to a point source applied to the plate in an arbitrary direction is modified in order to describe the case when the source is a seismic moment tensor of various tensor components. This part of the dissertation also discusses the analytical modeling of AE test sources such as pencil lead break, hammer hit excitation, etc.

In Part II of the dissertation, in situ experimental investigation and predictive modeling of AE from fatigue crack is discussed. The fatigue-crack growth-related AE needs to be separated from crack rubbing/clapping AE to understand the signals originating from the crack and to comprehend the situation of the crack in real-time. Novel SIF-controlled fatigue-crack growth experiment and vibration-induced crack rubbing/clapping experiment was invented for this purpose. The fatigue AE signals were recorded using piezoelectric wafer active sensors (PWAS). The AE source due to fatigue crack growth and rubbing/clapping was assumed as components of the moment tensor source discussed in Part-I, and the simulation was performed. The simulation results were compared with experimental observations, and a good agreement of simulation and experiment was observed.

Part III focuses on guided modeling of wave propagation in hollow cylinders. Normal mode expansion (NME) of hollow cylinder guided wave modes for a radial point source excitation is derived analytically. Mutually orthogonal guided wave modes in the hollow cylinder are used for NME, and the modal participation factors are determined analytically.

Part IV focuses on the experimental and analytical investigation of health monitoring of hollow cylinders using the passive and active health monitoring techniques discussed in Part III. Both passive and active structural health monitoring experiments are done on the TN32 dry cask storage scaled-down model. Then, the SAFE-NME method discussed in Part III was also used for the predictive simulation in a ‘6-inch schedule-40’ pipe.


© 2020, Roshan Prakash Joseph