We conducted deformation experiments in which temperature was increased above the thermal stability of antigorite to directly test the dehydration embrittlement hypothesis and its potential link to intermediate-depth earthquakes. The experiments were designed to maximize the possibility for instability: pore pressure was undrained, pre-dehydration stress was high, and deformation was localized prior to initiation of reaction. However, stick-slip instabilities (or unstable fault behavior) did not develop. Rather, we find that weakening occurs stably at a rate that depends on the ratio of the temperature ramp rate to the strain rate (), resulting in laboratory slow slip or stress relaxation events.

You do not currently have access to this article.