Nonlinear Viscoelastic Attenuation of Olivine Aggregates at High Stress Amplitudes

Linear viscoelasticity of olivine at low stress amplitudes has been studied extensively to examine attenuation of seismic waves. However, the effect of high stress amplitudes on attenuation has largely been ignored. Such high stress amplitudes are relevant to tidal dissipation in planetary systems with strong tidal interactions, such as Io or many identified exoplanets, and high stress amplitudes indirectly relate to the transient creep associated with large load changes, such as during postseismic creep or postglacial isostatic adjustment. To fill this gap in knowledge, we performed forced oscillation experiments on aggregates of San Carlos olivine with grain sizes of 78 m. Experiments were conducted in a Deformation-DIA apparatus coupled with synchrotron-based X-ray characterization at confining pressures of 35 GPa and temperatures of 12731473 K. We imposed oscillations in differential stress at frequencies in the range 0.33 mHz, amplitudes in the range 130210 MPa, and bias stresses less than 100 MPa. We observe strong attenuation (Q-1) of 0.1 to 2.4, which is considerably larger than attenuation observed in previous studies of olivine aggregates performed at lower stress amplitudes. This attenuation corresponds to a phase lag of 0.11.2 radians between stress and strain, with one experiment exhibiting a time delay of 670 s over a period of 3600 s. The viscoelastic response of olivine aggregates at these conditions is highly nonlinear, as revealed by a power-law dependence of attenuation on stress amplitude and cuspate hysteresis loops. Attenuation also increases with increasing bias stress, suggesting that dislocation density is a key microstructural variable. We suggest that dissipation of strain energy at high stress amplitudes is dominated by nonlinear viscoelasticity that can be attributed to the dynamics of lattice dislocations. Constitutive relations based on our data will lead to better extrapolation of existing flow laws and improved modeling of geodynamic processes with large stress changes.