https://doi.org/10.1130/G46516.1 Manuscript received 16 May 2019 Revised manuscript received 14 August 2019 Manuscript accepted 21 August 2019 © 2019 Geological Society of America. For permission to copy, contact [email protected]. Fault reactivation and strain partitioning across the brittle-ductile transition Gabriel G. Meyer, Nicolas Brantut, Thomas M. Mitchell and Philip G. Meredith Department of Earth Sciences, University College London, London WC1E 6BS, UK ABSTRACT ( localized-ductile transition) is not necessarily The so-called “brittle-ductile transition” is thought to be the strongest part of the litho- the same as the microscale transition in deforma- sphere, and defines the lower limit of the seismogenic zone. It is characterized not only by a tion mechanism (brittle-plastic transition) and transition from localized to distributed (ductile) deformation, but also by a gradual change that the two transitions can occur under differ- in microscale deformation mechanism, from microcracking to crystal plasticity. These two ent pressure and temperature conditions. The transitions can occur separately under different conditions. The threshold conditions bound- resulting complex interplay between brittle and ing the transitions are expected to control how deformation is partitioned between localized plastic mechanisms makes the flow stressσ flow fault slip and bulk ductile deformation. Here, we report results from triaxial deformation sensitive to a large number of parameters in the experiments on pre-faulted cores of Carrara marble over a range of confining pressures, ductile regime (see Evans and Kohlstedt, 1995, and determine the relative partitioning of the total deformation between bulk strain and and references therein), notably the imposed on-fault slip. We find that the transition initiates when fault strength σ( f) exceeds the yield strain rate and the accumulated strain. stress (σy) of the bulk rock, and terminates when it exceeds its ductile flow stress (σflow). In Furthermore, the criterion σflow > σf for the this domain, yield in the bulk rock occurs first, and fault slip is reactivated as a result of bulk onset of ductile deformation was originally es- strain hardening. The contribution of fault slip to the total deformation is proportional to tablished from studies on initially intact materials the ratio (σf − σy)/(σflow − σy). We propose an updated crustal strength profile extending the undergoing a simple monotonic loading history, localized-ductile transition toward shallower regions where the strength of the crust would and describes deformation regimes in a binary be limited by fault friction, but significant proportions of tectonic deformation could be ac- manner (localized or distributed) without empha- commodated simultaneously by distributed ductile flow. sizing the potential for coexistence of both fault slip and bulk ductile flow. The applicability of this INTRODUCTION AND distributed flow, regardless of the nature of the criterion to the crust might therefore be limited, METHODOLOGY deformation mechanism, and will use “brittle” to because crustal-scale deformation is controlled Under the low pressure and temperature con- describe fracturing processes at all scales. by preexisting structures (faults and shear zones; ditions of the upper crust, rocks generally deform In nature, the brittle-ductile transition zone see, e.g., Goetze and Evans, 1979; Brace and by grain-scale microcracking, and crustal-scale has been identified in exhumed shear zones Kohlstedt, 1980). Thus, it remains unclear if and deformation is accommodated by slip on discrete showing markers of crystal plasticity (e.g., my- how faults are reactivated across the brittle-duc- fault planes. In this regime, the overall strength of lonites) overprinted by slip planes and pseudo- tile transition. Previous experimental studies have the crust is limited by fault friction (Scholz, 2002; tachylytes that are inherent to the brittle regime commonly used sample geometries that enforce Paterson and Wong, 2005). Deeper in the crust, at (e.g., Sibson, 1980; Passchier, 1982; Hobbs sliding on narrow shear zones between essentially higher pressure and temperature, rock deformation et al., 1986). Such field evidence suggests that rigid blocks under increasing pressure and tem- becomes more diffuse, and may be driven by crys- the transition in deformation mechanism is as- perature conditions (e.g., Shimamoto, 1986; Pec tal plastic phenomena such as dislocation creep. sociated with a change in the degree of strain et al., 2016), which do not allow for quantification Here, the overall strength of rocks can generally localization, from narrow frictional slip zones of partitioning between fault slip and bulk strain. be described by a steady-state flow law sensitive to wider plastic shear zones. Here, we conducted rock deformation ex- primarily to temperature and strain rate (e.g., Goe- Laboratory experiments have shown that the periments on pre-faulted samples of Carrara tze and Brace, 1972; Evans and Kohlstedt, 1995). transition from localized fracture to ductile flow marble and monitored strain partitioning and The transition between these two rheological do- generally occurs when the frictional strength fault reactivation across the localized-ductile mains, the so-called “brittle-ductile transition”, of the fault, σf, equals the bulk flow stress of transition. Our experiments were performed at occurs over a pressure and temperature range the rock, σflow (Byerlee, 1968; Kohlstedt et al., room temperature and confining pressures (Pc) where rocks deform by an interplay of cracking 1995). However, distributed deformation at the from 5 to 80 MPa. We determined partitioning and crystal plasticity. The brittle-ductile transition macroscopic scale may still be dominated by of the total shortening between fault slip and commonly loosely refers to the progressive change brittle microscale processes, and only further off-fault matrix strain by subtracting the matrix in crustal rheology with increasing depth; here we increases in pressure and temperature lead to strain (measured with strain gauges) from the will use the term “ductile” in the sense described fully crystal-plastic flow. This shows that the total shortening (measured with external dis- by Rutter (1986), whereby it refers to macroscale macroscale transition in strain localization placement transducers). CITATION: Meyer, G.G., Brantut, N., Mitchell, T.M., and Meredith, P.G., 2019, Fault reactivation and strain partitioning across the brittle-ductile transition: Geology, v. 47, p. XXX–XXX, https://doi.org/10.1130/G46516.1 Geological Society of America | GEOLOGY | Volume XX | Number XX | www.gsapubs.org 1 Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/doi/10.1130/G46516.1/4850305/g46516.pdf by University College of London user on 17 October 2019 Experiments were conducted in two stages. RESULTS an additional experiment where a single, sec- During the first stage, samples were pre-faulted During the first stage Fig. 1( ), the sample ond-stage deformation cycle was performed at by loading at Pc = 5 MPa until localized brittle behaves in a manner typical of the brittle re- Pc = 35 MPa, which shows no further deviation failure occurred. Following failure, an additional gime, and the stress drop (accompanied by in matrix strain for a total shortening of up to increment of shortening ε˙ = ΔL/L (L—length) of partial relaxation of the off-fault elastic strain) a further 2% (Fig. DR4). Finally, at the highest either 0.1% or 1% was allowed to accumulate marks the formation of the macroscopic shear Pc (70 MPa and above), εmatrix remains equal to before proceeding to the second stage, in order fault. During the second stage, at each confining ΔL/L throughout the deformation cycle, which to test any effect of accumulated fault slip on the pressure step, the stress-shortening relationship implies that the fault is fully locked. transition. In the second stage, Pc was increased is initially linear, but deviates from linearity at To assess the extent of microcracking in the stepwise from 5 to 80 MPa in 5 or 10 MPa incre- some threshold stress σy, and then tends to pla- matrix, we measured the horizontal P-wave speed ments. At each pressure step, the samples were teau (Fig. 1A). This “plateau” stress increases across the fault during each deformation cycle −5 −1 reloaded at an axial shortening rate of ε˙ = 10 s significantly with increasingP c. At low Pc (10 (Fig. DR5). The wave speed at the start of each until 0.1% of irrecoverable axial shortening was and 20 MPa), the matrix strain (εmatrix) initially cycle increased with confining pressure. Dur- accumulated, and then unloaded before proceed- increases at the same rate as the total shortening ing deformation, the wave speed changed very ing to the next pressure step (see Section DR1 (ΔL/L), then deviates toward a constant value. little for cycles at Pc <30 MPa, but decreased and Fig. DR2 in the GSA Data Repository1 for The deviation point occurs at a stress denoted progressively for all cycles at higher pressures. an extended methodology, and Table DR3 for a σf, and marks the onset of fault slip (triangles The magnitude of the decrease in P-wave speed summary of experimental conditions). in Fig. 1B). At intermediate Pc (30–60 MPa), increased with increasing Pc from 30 to 60 MPa the same deviation is observed to occur, but but then decreased at higher confinement. εmatrix continues to increase beyond this point, At Pc = 10 and 20 MPa, the yield stress and albeit at a lower rate, indicating contributions the fault strength are equal, and the calculated 1GSA Data Repository item 2019391, extended methodology, supplemental figures, and data, is available from both matrix strain and fault slip to the to- slip contributes close to 100% of the total short- online at http://www.geosociety.org/datarepository/ tal shortening. This observation appears to be ening (Figs. 2A and 2D). Between Pc = 30 MPa 2019/, or on request from [email protected]. independent of shortening, as demonstrated in and Pc = 60 MPa, σf increases linearly with Pc, whereas σy remains approximately constant at A ∼115 MPa.