1 INTRODUCTION
The importance of joints in the attenuation of high frequency seismic waves originates from the fact that the mechanical properties of rock joints contrast greatly in comparison to the mechanical properties of intact rock. Efforts to incorporate fractures in high frequency wave propagation models have primarily been based on the model of Schoenberg [1980], where an individual joint is treated macroscopically as a displacement discontinuity [see Pyrak-Nolte et al., 1990]. When an elastic wave encounters a displacement discontinuity, part of the energy is reflected and part transmitted. The transmission coefficient is a function of the stiffness of the displacement discontinuity (the joint stiffness), the frequency of the elastic wave, and the acoustic impedance of the host material. While the displacement discontinuity appears to fit experimental data quite well at ultrasonic frequencies, questions arise due to the fact that the dynamic stiffnesses are consistently higher than quasi-static stiffnesses measured using the same sample.
In this study, joint closure and ultrasonic wave transmission experiments are conducted on an artificial joint in glass. Glass was chosen because it is a simple, homogeneous elastic material, a characteristic which greatly simplifies the task of separating the effects of deformation of the matrix from those of the joint. In previous work on natural joints in granite [Pyrak-Nolte et al., 1990], measurements based on ultrasonic transmission coefficients indicated that the dynamic stiffness was considerably greater than quasi-statically measured stiffnesses. They found compressional wave stiffnesses to be 2 to 5 times greater than quasi-static unloading stiffnesses, with a high degree of variation from one sample to the next, as well as from one normal load to the next. In addition, compressional stiffnesses were found to be greater than shear stiffnesses by as much as a factor of 5 (typically near 2), again with considerable variation from sample to sample.
