Near-surface fracture detection in structural elements, investigation using Rayleigh waves

Abstract

The increased use of wave-based nondestructive techniques in characterizing existing infrastructure is restricted by their ability to detect relevant structural conditions such as anomalies. Transmission and reflection measurements provide average internal information about the structure elements. Although transmission data can be tomographically inverted to determine internal material parameters, damage is usually initiated near the surface of the structural element. Rayleigh waves that propagate along the surface of an object are ideally suited for the detection of near surface defects.

This research investigates the use of Rayleigh waves for the identification of near surface fractures in structural elements. The study involves a conceptual analysis, finite element modeling and small scale experimentation. Initial work on thin Plexiglas sheets develops the methodology of Rayleigh wave measurement and examines the interaction of a Rayleigh wave with a slot. Subsequent finite element modeling further advances the understanding of the Rayleigh wave/fracture interaction. The final step uses the methodology and knowledge gained from Plexiglas plates to study the ability of Rayleigh waves to detect slots in small scale concrete beams.

To begin, the study focuses on the characteristics of a Rayleigh wave formed in an infinite half-space. A subsequent chapter introduces the signal processing techniques and algorithms used to measure Rayleigh wave dispersion and energy density in the frequency-wavenumber (FK) domain. Experimental measurements on two-dimensional Plexiglas analogues define the appropriate test procedures and interpretation criteria needed for the characterization of Rayleigh waves. In plates, Rayleigh waves form by the superposition of fundamental Lamb modes at high frequencies and wavenumbers. After establishing the existence of Rayleigh waves in the thin plate, time-acceleration measurements are made at different locations on the Plexiglas plate, with respect to a slot of varying depth. The slot effectively blocks wavelengths of the Rayleigh wave shorter than the slot depth. Frequency-wavenumber data show reflections of the Rayleigh wave from the front of the slot, where the strength of reflection increases as the slot depth increases.

Finite element modeling provides additional knowledge about the Rayleigh wave/fracture interaction. Initially, the finite element model is calibrated using experimental data and material parameters quoted in the literature. Subsequent simulations study time-acceleration measurements made at different locations inside the plate for various slot depths. The Rayleigh wave formed behind the slot is a combination of long wavelength energy passing the slot and short wavelength mode converted Lamb waves.

A series of experiments further examines the slot detection ability of Rayleigh waves in small concrete and cement beams. Initial measurements provide insight into Rayleigh wave motion at different locations on the beam. A finite element model calculates theoretical dispersion curves for comparison with experimental results. In addition, the finite element model illustrates that Rayleigh waves form by the superposition of fundamental flexural and longitudinal modes at high frequencies and wavenumbers. Preliminary measurements show that the longest wavelength of an ideal Rayleigh wave is 1/2 the beam thickness. A set of receiver array measurements examines the effect of a slot cut on Rayleigh wave for only the shortest slot depth. Rayleigh wave reflections are not strong enough to confirm the location of the slot. High material attenuation reduced the Rayleigh wave energy.

This method would benefit by additional work examining different receiver arrangements and frequency regulated input sources. Also, further theoretical and experimental work should focus on combining the knowledge acquired from Rayleigh waves with information gained from observed higher vibrational modes.