Progressive changes in rock fracture mechanical properties with exposure age at the Earth's surface.

Fractures in rocks are ubiquitous, from grain-scale microcracks to crustal scale faults. Importantly, fracture networks allow crystalline rocks to store and transport fluids, which can then interact with rock-forming minerals and enhance rock deformation processes, leading to slow, progressive fracture growth that is time-, stress- and environment-dependent.

 

We previously reported progressive changes in physical and mechanical properties of granitoid boulders exposed at the surface from zero to around 100 ka in Eastern California, USA. We noted systematic decreases in tensile strength, uniaxial compressive strength, elastic modulus and seismic wave velocities, and systematic increases in porosity and permeability, with increasing surface exposure age. We postulated that the observed changes were likely functions of an increase in the level of crack damage over time. Hence, we interpreted the changes as reflecting progressive subcritical crack growth arising from ubiquitous, but relatively low magnitude environmental stresses acting continuously on the boulders over the extensive periods of exposure.

 

Here, to avoid ambiguity in interpretation, we report direct measurements of key fracture mechanical properties made on samples from the same boulders. The critical stress intensity factor for dynamic fracture propagation (fracture toughness, K_IC) was measured using the Double Torsion testing methodology. We also measured the pre-exponential offset (A) and the subcritical crack growth index (n) in the Charles’ Law relation: V = A (K_I/K_IC)ⁿ, using the same technique (where V is the crack growth rate). We find that K_IC decreases from around 2.0 MPa.m^-1/2 in fresh material to around 0.5 MPa.m^-1/2 in boulders exposed for around 90 ka, and that the A offset increases from -5 to +20. By contrast, we find no significant change in the n index, which has a value of around 60 ± 10, apparently regardless of exposure age.

 

These results suggest that the ease of nucleation and rate of growth of new cracks increases with exposure age, consistent with rocks weakening over time through decreasing strength. However, this is in direct contradiction with extensive field measurements that appear to show that the rate of crack growth (measured by crack intensity on thousands of rocks) decreases over time.

 

So, how do we reconcile these apparently contradictory observations? Here we observe that over exposure time, individual cracks can nucleate and then continue to grow under low imposed stress. However, this assumes that the stress reaching the crack tip does not change over time. But, in nature, the stress intensity (K_I) felt at the crack tip is a function of the stress propagating throughout the rock mass. As the bulk rock becomes more compliant (lower Young’s modulus) over exposure time, due to diffuse microcracking, the rock accommodates more elastic strain and translates a lower net stress intensity to each individual crack tip. Thus, we expect the overall rate of cracking in any rock mass to depend on the instantaneous ratio between the decreasing stress intensity factor and the decreasing fracture toughness (i.e., K_I/K_IC).