Non-contact measurements to estimate the elastic properties of rocks under in situ conditions

Abstract

Laboratory rock physics measurements are important for understanding how the physical properties of rocks control the behaviour of elastic waves propagating in the earth. Traditionally, ultrasonic waves are excited and detected in the laboratory with contacting transducers inside fluid-filled pressure vessels that replicate in situ subsurface confining stress. Instead of transducers, we use laser ultrasonics (LUS) to generate and record ultrasound in rock samples, an entirely non-contact technique. This method offers several advantages over transducer measurements: mechanical coupling issues are avoided, very broadband (30 kHz to 24 MHz) waveforms can be recorded, the small footprint of the laser beams allows a single rock to be densely sampled, and group velocity is always measured. However, LUS has so far been limited to studies at atmospheric pressure, and since the elastic properties of rocks are strongly dependent on confining stress, the advantages of LUS have not yet been realised for realistic rock physics measurements under in situ conditions. We have designed and implemented a methodology to perform non-contact LUS compressional wave measurements under in situ confining stress whereby rock samples are mounted inside a pressure vessel with two optical windows for the source and receiver laser beams. Experimental acquisition and arrival time picking are both automated. To demonstrate the advantages of this technique, we investigated the anisotropy and pressure dependence of four rocks from the Alpine Fault in New Zealand. Due to the dense sampling, we experimentally determined the orientation of transverse isotropy for three protolith samples and showed that transverse isotropy was not a valid assumption for a highly fractured cataclasite sample. Fitting a curve to over 90 independently measured Pwave velocities for each sample significantly improved both the accuracy and precision in the estimates of the elastic constant c13. Although the rocks had similar mineralogy, the observed differences in velocity, anisotropy, attenuation, and pressure dependence were mainly controlled by variations in microcrack density and alignment at shallow crustal pressures. In the future, we intend to improve this methodology to exploit the advantages of noncontact LUS measurements under in situ conditions for a range of rock, ice, and material physics experiments.