Rocks commonly display a reduction in elastic modulus during an acoustic perturbation followed by a logarithmic recovery of the modulus to its original value. This nonlinear elastic behavior can also be induced by changes in temperature of the rock, although experimental data for this are sparse. In this paper, we utilize temperature perturbations to broaden our understanding of the causes and mechanisms of nonlinear elasticity in rocks. We perform laboratory experiments tracking the nonlinear response of a Berea sandstone under vacuum both during and following 10°C changes in temperature. Ultrasonic velocity decreases by up to 0.30 ± 0.02% during both temperature increases and decreases, before recovering in the following days. Continuous monitoring over a 2‐month period also reveals a longer‐term 0.78 ± 0.04% increase in velocity. To explain these observations, we modify a recent model for nonlinear elasticity based on the shearing and subsequent recreation of internal microscopic contacts. Application of this model to our data suggests that internal stresses from differential thermal expansion/contraction of mineral grains break contacts, inducing nonlinear weakening. The degree of weakening depends on the temperature gradient. Slow dynamics recovery in the ∼10 hr following a temperature change likely results from the recreation of broken contacts due to nanoscopic adhesive forces. In contrast, the long‐term velocity increase results from evaporation of bound water from clay minerals. In addition to furthering our understanding of nonlinear elasticity in rocks, our results imply that sudden changes of temperature in shallow crustal environments will induce nonlinear weakening that persists for hours to days.

SUMMARY Dynamic nonlinear elasticity in rocks may play an important role in earth processes, such as earthquake nucleation. In order to understand how nonlinear elasticity occurs within the shallow crust, experiments are required that simulate the in situ conditions of intact crustal rocks. Additionally, exploring the behaviour of nonlinear elasticity in response to changes in external parameters (e.g. temperature and wave frequency) acts as a means to further illuminate the complex mechanisms which give rise to nonlinear elasticity in rocks. In this study, we perform dynamic acoustoelastic testing (DAET) experiments on an intact cataclasite from the damage zone of the Alpine Fault, New Zealand. By performing pump-probe DAET experiments inside a temperature-controlled chamber, we are able to investigate a rich variety of nonlinear behaviour as a function of temperature. We find that the magnitude of average softening, cubic nonlinearity, and hysteresis tend to increase as temperature increases from 20 to 110 °C. In contrast, quadratic nonlinearity decreases with increasing temperature. These observations support the hypothesis that at least two distinct mechanisms control nonlinear phenomena in rocks. Nonlinear parameters show little to no dependence on frequency over the 200–600 Hz pump range, although values of the nonlinear parameter α are found to be nearly two orders of magnitude smaller than those determined using ultrasonic perturbations. Additionally, an analysis using different time windows shows that the surface waves of the ultrasonic probe sense greater nonlinearity compared to the direct P- wave due to differences in the polarization and propagation paths. As well as providing further insight into the mechanisms responsible for nonlinear elasticity in rocks, our experiments show that nonlinear softening will increase as temperature increases in the damage zones of faults. This has potential implications for understanding earthquake nucleation processes.

We perform laser ultrasonic measurements to investigate the spatial dependence of dynamic nonlinear weakening in rocks from the Alpine Fault, New Zealand. Rocks outside the damage zone display no nonlinear weakening. Within the damage zone (<30 m from the fault), cataclasites present a 3% reduction in shear modulus from wave amplitudes inducing 1–2 microstrain at atmospheric pressure. This nonlinear elasticity decreases with a characteristic pressure between 1 and 2.5 MPa. We show that rock weakening is therefore strongest in the near surface. However, this significant elastic nonlinearity in cataclasites at low strains confirms that rock weakening may play an important role in earthquake processes, such as fault weakening, triggering of slip, rupture propagation, and coseismic velocity decreases.

P-wave anisotropy is significant in the mylonitic Alpine Fault shear zone. Mineral- and texture-induced anisotropy are dominant in these rocks but further complicated by the presence of fractures. Electron back-scattered diffraction and synchrotron X-ray microtomography (micro-CT) data are acquired on exhumed schist, protomylonite, mylonite, and ultramylonite samples to quantify mineral phases, crystal preferred orientations, microfractures, and porosity. The samples are composed of quartz, plagioclase, mica and accessory garnet, and contain 3–5% porosity. Based on the micro-CT data, the representative pore shape has an aspect ratio of 5:2:1. Two numerical models are compared to calculate the velocity of fractured rocks: a 2D wave propagation model, and a differential effective medium model (3D). The results from both models have comparable pore-free fast and slow velocities of 6.5 and 5.5 km/s, respectively. Introducing 5% fractures with 5:2:1 aspect ratio, oriented with the longest axes parallel to foliation decreases these velocities to 6.3 and 5.0 km/s, respectively. Adding both randomly oriented and foliation-parallel fractures hinders the anisotropy increase with fracture volume. The anisotropy becomes independent of porosity when 80% of fractures are randomly oriented. Modeled anisotropy in 2D and 3D are different for similar fracture aspect ratios, being 30 and 15%, respectively. This discrepancy is the result of the underlying assumptions and limitations. Our numerical results explain the effects that fracture orientations and shapes have on previously published field- and laboratory-based studies. Through this numerical study, we show how mica-dominated, pore-free P-wave anisotropy compares to that of fracture volume, shape and orientation for protolith and shear zone rocks of the Alpine Fault.

Quantifying the amount and alignment of microfractures is important to understand the geomechanics, fluid flow, and seismic imaging of fault zones. At the Alpine Fault, New Zealand, the preferred alignment of minerals, foliation, and fractures results in elastic wave anisotropy. We have designed a unique laser‐ultrasonic laboratory setup to study Alpine Fault rock samples at upper crustal conditions. Combined with differential effective medium modeling, we distinguish microfracture porosity and orientation from mineral alignment, as a function of distance to the principal slip zone (PSZ). Nearest to the PSZ, the cataclasite has the lowest P wave anisotropy with the most (randomly oriented) fractures. Next, the ultramylonite exhibits the greatest P wave anisotropy (∼45%) with 40% of its fractures aligned with foliation. Further from the PSZ, P wave anisotropy is 14–19% on average, due to 20–30% of the fractures being oriented in the same direction as mineral alignment.

We present a new noncontact methodology to excite and detect ultrasonic waves in rocks under in situ pressure and temperature conditions. Optical windows in the side of a pressure vessel allow the passage of a laser source and a receiver for noncontact laser ultrasonic measurements. A heating mantle controls the temperature, and a rotational stage inside the vessel makes it possible to obtain measurements as a function of angle. This methodology is the first to combine the advantages of laser ultrasonics (LUS) over traditional transducer methods with measurements under in situ pressure and temperature conditions. These advantages include the absence of mechanical coupling, small sampling area, and broadband recordings of absolute displacement. After describing the experimental setup, we present control experiments to validate the accuracy of this new system for acquiring rock physics data. Densely sampled rotational scans performed on an Alpine Fault ultramylonite rock reveal a decrease in P-wave anisotropy from 62% at atmospheric pressure to 36% at 16 MPa. This result highlights the importance of performing rock physics measurements under in situ confining stress and demonstrates the advantages of the methodology for investigating anisotropy. In addition, a 5.6% decrease in the P-wave velocity of the ultramylonite sample between 20 °C and 100 °C at a constant 10 MPa confining stress demonstrates the capability of this new methodology for acquiring data under both in situ pressure and temperature conditions. This new methodology opens the door for probing the pressure and temperature dependence of the elastic properties of rocks and other materials using LUS techniques.

The physical properties of wood are of interest in a wide range of applications, including biology, climate science, and civil engineering. Its orthotropic nature poses considerable challenges in assessing the fully three-dimensional properties of wood. Laser ultrasonic waveforms of Radiata Pine of differing quality illustrate opportunities to extract its angle-dependent physical properties. We present a newly-derived Modified Hankinson Equation that improves the fit to the observed angle-dependence of wave speed in the higher quality sample by almost a factor of 3 to the original Hankinson Equation. Wave forms in the lower quality sample detect small-scale heterogeneity (due to pith, in this case), based on variations in ultrasonic wave speed and attenuation.

Education in physics and geosciences can be effectively illustrated by the analysis of earthquakes and the subsequent propagation of seismic waves in the Earth. Educational seismology has matured to a level where both the hard-and software are robust and user friendly. This has resulted in successful implementation of educational networks around the world. Seismic data recorded by students are of such quality that these can be used in classic earthquake location exercises, for example. But even ocean waves weakly coupled into the Earth's crust can now be recorded on educational seismometers. These signals are not just noise, but form the basis of more recent developments in seismology, such as seismic interferometry, where seismic waves generated by ocean waves-instead of earthquakes-can be used to infer information about the Earth's interior. Here, we introduce an earthquake location exercise and an analysis of ambient seismic noise, and present examples. Data are provided, and all needed software is freely available.