Is frictional heating needed to cause dramatic weakening of nanoparticle gouge during seismic slip? Insights from friction experiments with variable thermal evolutions

To examine whether faults can be lubricated by preexisting and newly formed nanoparticles, we perform high‐velocity friction experiments on periclase (MgO) nanoparticles and on bare surfaces of Carrara marble cylinders/slices, respectively. Variable temperature conditions were simulated by using host blocks of different thermal conductivities. When temperature rises are relatively low, we observe high friction in nano‐MgO tests and unexpected slip strengthening following initial weakening in marble slice tests, suggesting that the dominant weakening mechanisms are of thermal origin. Solely the rolling of nanoparticles without significant temperature rise is insufficient to cause dynamic fault weakening. For nano‐MgO experiments, comprehensive investigations suggest that flash heating is the most likely weakening mechanism. In marble experiments, flash heating controls the unique evolutions of friction, and the competition between bulk temperature rise and wear‐induced changes of asperity contact numbers seems to strongly affect the efficiency of flash heating. Key Points The presence of nanoparticles alone does not cause weakening without sufficient temperature rise Flash heating is the most likely weakening mechanism in experiments on nano‐MgO gouges Bulk temperature rise and change in contact numbers affect flash weakening on marble bare surface