ABSTRACT:
The low frictional resistance to rock-on-rock sliding reported in large blockslides, in coseismic fault rupture and in laboratory-scale rock friction tests has been attributed to a variety of causes. Herein we propose a mechanical explanation for the reduced friction, which seems likely to be universally relevant to complement other mechanisms. Rock-on-rock sliding of intact brittle rocks always generates a layer of comminuted debris. Rock must fail in order to form and further comminute debris; at local strain rates >> 100 s-1, recycling of elastic strain energy stored in accomplishing fragmentation generates instantaneous, local, GParange isotropic pressures similar to the rock’s Hugoniot elastic limit (Q). Under rapid strain, simultaneously fragmenting grains deliver large normal forces to the boundaries of the comminuting layer, reducing the confining stress on the debris (and hence its resistance to shear), thus lowering the frictional resistance to slip. This behaviour corresponds quantitatively to published laboratory data on granite friction; to the dynamics of low-angle blocksliding and faulting; and to our data on rapid shearing of fragmenting dry coal.
1 INTRODUCTION
Anomalously low frictional resistance to the motion of rock masses has been reported or inferred from time to time in a number of geological situations. Lower-than-normal resistance to rock-on-rock slip is widely inferred in fault motion (e.g. Scholz 2002, Parsons 2002, Townend & Zoback 2004), which involves shear in comminuted rock debris. Rapid, lowangle motion of large intact blockslides generates comminuted material at the sliding surface (Anders et al. 2000) and unequivocally requires low basal friction (Davies et al. 2006 For example, Di Toro etal. (2004) attribute their dramatically low friction data to formation of weak silica gel in the presence of moisture on the laboratory rock surface, which is difficult to apply to the much larger-scale situation of blocksliding.
