Structural Health Monitoring (SHM) and Condition Based Maintenance (CBM) systems can provide substantial benefits for aging aerospace systems as well as newer systems still in the design process. In aging aerospace systems, a retrofitted SHM system would alert users of incipient damage preventing catastrophic failure. For newer systems, incorporating a SHM approach and using CBM techniques can reduce life-cycle costs. Central to such SHM and CBM systems is the ability to detect damage in a structure.

Traditional approaches to damage detection in structures involve one of two methods. In the modal dynamics approach, the natural frequencies and modeshapes of a structure shift when damage occurs. The location, type, and amount of damage is determined by the shifts in the modal properties due to damage. Alternately, in an Ultrasonics approach, the structure is scanned with a specialized transducer which induces high frequency vibrations in the structure. Damage in the structure is inferred when these vibrations are altered. In the same vein as Ultrasonics, Acoustic Emission based methods listen for energy release in the structure upon defect growth. All of these techniques have limitations which hinder their usage in a practical system. This thesis attempts to develop a methodology with the benefits of the modal approach as well as the Ultrasonics/Acoustic Emission approach. The methodology is commonly referred to as an Acousto-Ultrasonic technique for damage detection.

The structural dynamics of plate structures is described as wavelike in nature where the plate is a medium for wave propagation. For thin plates, bulk wave propagation is described using Lamb wave modes. The two fundamental modes of wave propagation are the in-plane acoustic mode and the transverse bending mode. The interaction of these waves with a discontinuity or damaged region changes the way the waves propagate. Part of the incident wavefront is reflected back while the rest is transmitted through the damaged region.
The presence of the reflected waves and the attenuation of the incident wave in the transmission case indicates that damage is present in the structure.

A reflected wave can be used to infer the location of damage on a structure. A phased array technique is used which isolates the source of these reflections in an effort to identify the damage location. Phased arrays are shown to act as directional filters which sense the structural vibrations of the plate and selectively look in a certain direction on the plate relative to the array orientation. Reflections from damage can then be extracted from sensor signals in an effort to locate a damage region. Damage in the form of a hole in an isotropic plate is examined and the phased array technique is used to show that the location of the hole can be determined as well as trends indicating that the size can be determined.

For anisotropic materials, such as fiber reinforced composite laminates, the dynamics associated with wave propagation are more complicated. This is shown in composite laminates where wave propagation is dependent on the direction of travel. A more fundamental concern is developing an understanding of the wave propagation properties of a composite laminate. A model is developed to determine the dispersion relations for a laminate made of an arbitrary layup of orthotropic plies. The predicted results from the model are compared with experimental results showing that the dispersion relations can be accurately determined.

The phased array technique is then applied to wave propagation in a composite laminate to determine the presence of delamination damage. It is shown that delaminations do produce reflections from incident waves. Though the reflections are weak, they can be extracted from the structural dynamics by use of the phased array technique.

Traditional modal methods are then compared with a wave propagation approach. Modal methods are shown to have difficulty ascertaining low damage levels when using a small number of modes. As the amount of information increases, such as a larger set of modes and larger number of sensors, the modal methods become more sensitive. A wave propagation approach is shown to be sensitive to small damage amounts and the location and approximate extent of damage can be determined using less information.