We have conducted ultrasonic experiments, between 0.3 and 1 MHz, to measure velocity and attenuation ($Q^{-1}$) anisotropy of P- and S-waves in dry Whitby Mudstone samples as a function of stress. We found the degree of anisotropy to be as large as 70% for velocity and attenuation. The sensitivity of P-wave anisotropy change with applied stress is more conspicuous than for S-waves. The closure of large aspect-ratio pores (and/or micro cracks) seems to be a dominant mechanism controlling the change of anisotropy. Generally, the highest attenuation is perceived for samples that have their bed layering perpendicular (${90}^{°}$) to the wave path. The observed attenuation in the samples is partly due to the scattering on the different layers, and it is partly due to the intrinsic attenuation. Changes in attenuation due to crack closure during the loading stage of the experiment are an indication of the intrinsic attenuation. The remaining attenuation can then be attributed to the layer scattering. Finally, the changes in attenuation anisotropy with applied stress are more dynamic with respect to changes in velocity anisotropy, supporting the validity of a higher sensitivity of attenuation to rock property changes.

The transition from recoverable elastic to permanent inelastic deformation is marked by the onset of fracturing in the brittle field. Detection of this transition in materials is crucial to predict imminent failure/fracturing. We have used an ultrasonic pulse transmission method to record the change in waveform across this transition during fracturing experiments. The transition from elastic to inelastic deformation coincides with a minimum in ultrasonic attenuation (i.e., maximum wave amplitude). Prior to this attenuation minimum, the existing microfractures close. After this minimum, new microfractures form and attenuation increases until peak stress conditions, at which point, larger fractures form leading to complete sample failure. In our experiments, velocity changes are not sensitive enough to be indicative for the transition from elastic to inelastic deformation. Analysis of attenuation, not velocity, may thus detect imminent failure in materials. Our results may help detect fracturing in borehole casings or the near-wellbore area, or they may help predict imminent release of energy by seismic rupture.