of 20
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
2
Fa
ult
rock
heterogeneity
can
produce fault
weakness and
reduce fault stability
1
2
John D. Bedford
1
*
, Daniel R. Faulkner
1
& Nadia Lapusta
2
,3
3
1
Rock Deformation
Laboratory, Department of Earth, Ocean and Ecological Sciences, University of
4
Liverpool
5
2
Department of
Mechanical and Civil Engineering, Division of Engineering and Applied Science,
6
California Institute of Technology
7
3
Seismological Laboratory,
Division of
Geological and Planetary Sciences, California Institute of
8
Technology
9
*Corresponding author: jbedford@liverpool.ac.uk
10
11
Abstract
12
Geological heterogeneity is abundant in crustal fault zones
;
h
owever,
its
role in controlling the
13
mechanical
behaviour
of faults is poorly constrained. Here,
we present
laboratory friction experiments
14
on
laterally heterogeneous faults,
with
patches of strong, rate
-
weakening quartz
gouge
and weak, rate
-
15
strengthening clay gouge
.
T
he experiments show
that the
heterogeneity leads to
a significant reduction
16
in
strength
and
frictional stability in comparison to compositionally identical faults
with
homogeneously
17
mixed
gouges
.
We identify
a combination of weakening effects
, including
sm
earing of the weak clay;
18
differential compaction of the two gouges redistributing normal stress; and shear localization producing
19
stress concentrations in the strong quartz patches. The results demonstrate
that
geological heterogeneity
20
and its evolution can
ha
ve
pronounced effects on
fault
strength
and stability
and
,
by extension
,
on
the
21
occurrence of slow
-
slip transients versus earthquake
rupture
s
and the characteristics of the resulting
22
events, and
should
be
fu
rther studied
in lab experiments and earthquake source model
l
ing
.
23
24
25
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
3
Introduction
1
M
any
large
crustal faults have been shown to
be
frictional
ly
weak
1
6
when compared to
2
laboratory measurements of
quasi
-
static
fault friction. T
he
coefficient of friction
=
̅
̅
̅
, where
is
3
the shear stress
during slip
and
̅
̅
̅
is the effective normal stress, of most
geological
materials is typically
4
measured in the laboratory to be between
0.6
-
0.8
5
at slow slip
speeds
,
independent of rock type
7
, with
5
the exception of a few weak minerals,
predominantly
phyllosilicates
7,8
.
Possible explanations for weak
6
faults in nature
, where
the
apparent
at which faults operate
is
often
less than
0.5,
include localization
7
of weak minerals along structural foliations
9
13
,
dynamic weakening during seismic slip
14
,
and elevated
8
pore fluid pressure
interpreted as lower friction coefficient
s
15,16
.
As well as being apparently weak,
9
many crustal faults also exhibit a
spectrum
of slip behaviour
, with earthquake slip and aseismic creep
10
often
occurring on the same fault
17,18
and slow slip phenomena being prevalent at all crustal depths
19
.
11
While
the appa
rent weakness of faults and spectrum of slip behaviour
can be
attributed to
the
effects of
12
spatially varying and temporally evolving confinement, temperatur
e
, and pore fluid pressure
, it is
clear
13
that heterogeneity
in fault
zone rocks
(Fig. 1a)
can also play a
n important
20,21
,
if not domina
nt,
role
.
14
Geological investigations have shown that heterogeneity
in fault
z
one rocks
occurs over many
15
different scales, from
submillimetre
-
scale
structural foliations
9,10
,
centimetre
-
to
meter
-
scale
blocks
16
with
in a
shear zone
mélange
22
,
hundred
s
-
of
-
meter
s
scale
where lens
es
of damaged protolith
can be
17
entrapped within the core of wide (km
-
scale) fault zones
23,24
(e.g. Fig. 1a)
,
to
tens
-
of
-
kilometers
scale
18
variations in
rock types
10,25
.
The role of large
-
scale fault
rock
heterogeneity has been highlighted in a
19
number of studies; for example, i
t has be
en
suggested
that
heterogeneities such as
seamounts
can
act as
20
earthquake
nucleation sites and
control the
seismogenic
behaviour of
subduction
zone
megathrust
21
faults
26,27
.
However, the
importance
of
small
-
scale
fault
rock heterogeneity
in controlling fault slip
22
behaviour
,
average
fault strength
,
and fault stability
is
still uncertain
.
23
Here
,
the
effect
of
fault rock
heterogeneity on fault strength and slip behaviour
is investigated
24
by a series of laboratory friction experiments on simulated laterally heterogeneous faults
.
The faults
25
consist of
different sized
patches of strong
, rate
-
weakening
quartz
,
and we
ak
, rate
-
strengthening
clay
26
fault gouge
s
.
Until now,
the majority of
experimental investigations
have been performed
using
27
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
4
mixtures of different fault gouge materials with varying frictional properties
,
where the materials are
1
homogeneously mixed together
28
31
;
intact wafers of natural gouge
have also been
used
9,10
.
In this work,
2
experiments are performed on both homogeneous
ly mixed
and
spatially
heter
ogeneous gouge layers
3
consisting of quartz, frictionally strong and rate
-
weakening, and kaolinite clay powder, frictionally
4
weak and rate
-
strengthening. The fault gouge layers
(50 mm long, 20 mm wide, with
a
thickness of
~
1
5
mm
at the onset of shear, after
initial pressurization
)
are sheared in
a direct shear arrangement (
Fig. 1b,
6
see also
Supplementary
Fig. 1
) within a triaxial deformation apparatus (see
M
ethods). The
7
heterogeneous
gouge layers
are constructed
by placing different sized patches of
fine
-
grained
quartz
8
and clay
powder (both <5 μm grain size)
adjacent to each other
in a symmetrical pattern, with a
central
9
quartz patch being bound by two clay patches (Fig. 1
b
)
. This symmetrical arrangement ensure
s
t
hat no
10
misalignment between the direct shear forcing blocks
would occur as a
result
of
any
differential
11
compaction between the different materials
;
furthermore,
the amount of gouge material used (measured
12
by weight prior to the experiment) was calculated s
o that the thickness of the quartz and clay gouge
s
13
were the same after
initial pressurization and
a small amount of shear (
Supplementary
Fig. 2
).
The
14
normal stress is applied by the confining pressure (
)
in the triaxial apparatus
, held constant at 60
MPa
15
for all tests in this study, and the pore fluid pressure (
)
within the gouge
is servo
-
controlled
at a
16
constant value of 20 MPa,
resulting in the effective normal stress
̅
̅
̅
=
40 MPa (
̅
̅
̅
=
)
.
The
17
gouge layer
s are
sheared
up to a maximum di
splacement of 8.5 mm
(shear strain ≈
10,
given
the final
18
layer thickness
of ~0.85 mm
)
. Monitoring t
he evolution
of
shear stress
while
applying
velocity steps
19
from
0.3
to
3
μm
·
s
-
1
and back
allows the experiments to
quantify
the rate
-
and
-
state
friction parameters
20
that determine the stability of fault slip
32
.
These sliding velocities are sufficiently slow
, given the gouge
21
permeability,
to ensure that pore pressure transients do not build up within the gouge layer
during
22
shearing
33
.
The size
s
of the strong
yet unstable
quartz and weak
but stable
clay patches
are
varied to
23
investigate the role of different scales of heterogeneity on
the magnitude and stability of
fault friction.
24
Results
25
The experimental results indicate pronounced differences between the behaviour of laterally
26
heterogeneous faults compared to the
laterally homogeneous faults with mixed gouge
(
F
igure 1
)
.
All
27
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
5
experiments are characterized by
an initially
rapid increase in shear stress during
the
loading
phase
,
1
before the samples
clearly
yield
-
i.e.,
shear inelastically
-
after approximately 1 mm of displacement.
2
After that
,
the friction coefficient
of
the
homogeneous
ly
mixed gouge layers
remains relatively
3
constant
(Fig. 1
d
)
,
with rate
-
and
-
state effects
consistent with results from previous experimental
4
studies
29
31
. In contrast
,
the heterogeneous gouge layers all show ubiqui
tous weakening
(Fig. 1
c
)
, with
5
evolving towards the value
of the weaker clay phase.
To ensure
that
the observed
weakening
was
not
6
caused by the
arrangement of
the different gouge patches
in the experiments
,
tests were performed
7
where the symmetry of the heterogeneous layers was reverse
d (
i.e.,
a central clay patch bound by two
8
quartz patches)
. These tests also exhibit similar weakening (
Supplementary
Fig. 3
) suggesting that it is
9
the heterogeneity itself, not the arrangement of the different materials,
that
causes the weakening.
Stable
10
sliding is observed for all
homogeneous
ly mixed
faults
and the majority of heterogeneous
faults
.
11
However
,
when
the quartz pa
tch
in the heterogeneous layers
comprises ≥80% of the total sliding area
,
12
unstable stick
-
slip sliding
emerges
, typically triggered by up
-
steps in the sliding velocity
(Fig. 1
c
)
.
13
The observed weakening of the heterogeneous faults is
greater
than
can be explained by the
14
observed smearing
of the clay patches.
Microstructural analysis of a heterogeneous layer recovered at
15
the end of an experiment (Fig.
2
a) shows smearing of clay into localized boundary Y
-
shears that
16
propagate into the quartz patch. W
ith progressive smearing and localization of the clay phase
(Fig. 2
b),
17
the strength of the layer overall is expected to decrease as a greater proportion of
the
slip
ping surface
18
can be
located within
the weak
clay
phase
34
. As the frictional strength of the endmember gouge
19
compositions is known (i.e. 100% quartz and 100% clay in Fig. 1
c
), th
e
predicted weakening
due t
o
20
smearing can be
calculated
(Fig. 2c)
by assuming
that
the overall strength is determined by the strength
21
of the two gouges acting in series, based on their relative proportions (the arithmetic mean of
,
based
22
on the proportions of clay and quartz withi
n the layer).
The predicted weakening, associated with the
23
relative increase of the clay patches is considerably less than the observed weakening in the experiments
24
(Fig. 2c), suggesting that clay smearing alone is not responsible for the progressive weake
ning of
25
heterogeneous faults.
26
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
6
An
additional
cause of the weakening could be differential compaction between the different
1
gouge materials resulting in a redistribution of normal stress
(see Supplementary Information for full
2
discussion of this effect)
.
The volumetric strain data from the endmember
quartz and clay gouge
3
experiments show that the quartz gouge experiences a greater layer thickness reduction of about
20
m
4
than the clay gouge
during slip
(Supplementary Fig. 2). In the heterogeneous layer experiments this
5
would result in
an increase of n
ormal stress on the weaker clay patches leading to a progressive
6
reduction in shear resistance, as observed in our experiments. Th
e
magnitude of this
effect is dependent
7
on the bulk (
) and shear (
) modul
i
35
, which
are
poorly constrained for the
gouge
materials in
this
8
study.
Using plausible values for the moduli (Supplementary Information) indicates that
this d
ifferential
9
compaction
effect
could potentially explain
a large component of the
weakening we observe in our
10
experiments
(Fig. 2d)
.
11
The velocity steps
from Figure 1
c
-
d
are
used to calculate the
evolution in the
rate
-
and
-
state
12
friction
36
38
paramete
r
(
a
b)
,
which
determine
s
the frictional stability of the fault
39
41
. W
hen
(
a
b
)
>
13
0
,
the sliding behaviour
is
rate
-
strengthening,
suppress
ing
instabilities
and promoting
stable sliding,
14
whereas when
(
a
b
)
<
0
,
the sliding behaviour is rate
-
weakening
which
promote
s unstable slip
15
behaviour and
the occurrence of stick
-
slip
s in the laboratory.
The
values of (
a
b
)
are consistently
16
lower
(i.e.
,
less
rate
-
strengthening
)
in
the
heterogeneous faults
throughout the experiment
s
(Figure 3a
-
17
c
)
.
This finding
indicat
es
that
the
heterogeneous faults are
closer to
the
potential
ly u
nstab
le, rate
-
18
weakening regime
than their homogeneous
counterparts
.
19
For
the homogeneous faults with
the pure quartz gouge and the
heterogeneous faults where the
20
quartz patch comprises ≥80% of the total sliding area, only the first velocity step can be
used to
21
determine the rate dependence
due to the occurrence of stick
-
slip instabilities triggered by subsequent
22
velocity steps.
Howeve
r
,
this initial velocity step at 1.5 mm displacement does show negative
values of
23
(
a
b
)
associated with
rate
-
weakening behaviour (Fig. 3
a), which is consistent with the occurrence of
24
stick
-
slip instabilities later in the experiment.
All of the calculated rate
-
and state friction data
are
25
presented in
Supplementary
Table 1
.
26
27
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
7
Discussion
1
Our experiments
show that
laterally
heterogeneous fault gouge layers weaken significantly in
2
co
mparison to homogeneous layers
, pointing to heterogeneity
-
induced weakening effects. We
3
hypothesize that the weakening occurs due to a combination of mechanisms, all of which can affect
4
natural fau
lts. The
mechanical smearing of the weak phase
with slip can reduce the overall shear
5
resistance as shear is likely to localize within the weak phase
34
, although
this mechanism by itself can
6
explain only part of the observed weakening
(Fig.
2
c
, see also Supplementary Fig. 4
)
.
Ano
ther
7
contributing mechanism
can be
d
ifferential compaction
of the weak and strong phases during shear
8
(
Supplementary
Fig. 2) which
would
result in a redistribution of normal stress along the
shearing layer,
9
with
the weaker phase supporting higher normal st
resses (
see Supplementary Information for further
10
discussion of this effect
).
The differential compaction can produce significant
weakening
effects
(Fig.
11
2d)
but it is poorly constrained
, with the conclusions based on end
-
member tests
of pure quartz and c
lay
12
samples
under constant normal stress
,
highlighting the need to better capture and characterize the
13
compaction/dilation effects in gouge experiments.
Finally,
additional weakening
can
be
due to
shear
14
occurring in the weaker
clay
gou
ge that produces
stress concentrations
along
localized
Y
-
shear bands
15
that propagate through
the stronger quartz patches
leading to enhanced weakening.
Similar shear stress
16
concentrations have also been suggested
to promote slip events in strong, rate
-
weake
ning gouge patches
17
in recent low normal stress experiments on decimeter
-
scale
heterogeneous
fault
s
42
.
Due
to difficulty
18
keeping the gouge layer intact during recovery at the end of our experiments, we were unable to acquire
19
detailed microstructural images of the tips of the propagating shear bands to look for evidence of
20
shear/damage zones in the quartz patch
. We do, however, observe
R
1
Riedel shears in the quartz patch
21
(Fig. 2a) which may help facilitate weakening by connecting the smeared clay on opposite sides of the
22
layer.
23
Competency contrasts between strong and weak materials in shear zone
mélanges
have b
een
24
s
uggested
previously to be important in controlling the
average
fault strength
and rheology
43
,
with only
25
a small amount of well
-
connected weak mater
ial needed to reduce fault strength
when structural
26
foliations are well developed
9
.
In our experiments, if the gouge layers could be taken to greater shear
27
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
8
displacements, the
clay smearing we observe along the edges of the quartz patch (Fig. 2a) would
1
ultimately
form a through
-
going layer of interconnected weak material
after a few centimetres of slip
.
2
Previous work has shown that such through
-
going layers can lead to a
reduction in the frictional strength
3
at slow slip velocities
11
and also
increase the
efficie
n
cy of
dynamic weakening at seismic slip velocities
4
(1
m
/
s
)
44
.
Although weak phase smearing
would
, to some extent, homo
genize the fault in the
overall
5
direction of shear, heterogeneity
would
likely always
be prevalent in natural faults, particularly
6
perpendicular to the slip direction and also at scales larger than investigated in this study
, as observed
7
in natural fault z
ones
25,45
.
Our results
s
how that the
average
frictional
strength of
laterally
heterogeneous
8
fault
s
is not just an average
of the respective
friction
properties
(Fig. 2c)
,
and that competency contrasts
9
can
substantially
reduce the fault strength,
even when
structural foliations are in their infancy
and
10
unconnect
ed
(Fig.
2
a
)
.
They also highlight the need to investigate further how different types of fault
11
heterogeneity, including
fault
-
parallel
and fault
-
normal heterogeneity,
and its evolution,
af
fect the
12
frictional behaviour
of faults
.
13
Contrasting material prope
rties within fault zones ha
ve
also been suggested to give rise to
14
mixed fault slip behaviour
46
and exert an important c
ontrol on earthquake rupture dynamics
47,48
.
15
Heterogeneities are also t
hought to strongly influence the sliding behaviour of other types of frictional
16
interface, such as at the base of glaciers
49
.
Our experiments show
that
heterogeneity
produces
an overall
17
reduction in stability
when compared to ho
mogeneous faults (Fig. 3
)
.
I
t should be noted
that a sufficient
18
amount of rate
-
weakening material is
still
required to p
romote
unstable slip
.
In our experiments, w
hen
19
the proportion of
the
rate
-
weakening material is
≤70%
, the heterogeneous faults
are stable overall, with
20
positive
(
a
b)
values
, although the values are closer to zero (and hence rate
-
neutral
behaviour)
than
21
those of
their homogeneous counterparts (Fig. 3)
;
however the behaviour remains rate
-
strengthening,
22
instabilities do not initiate and aseismic slip prevails.
Only
when the strong
rate
-
weakening
patch
23
comprises ≥80% of the layer
do
stick
-
sli
p instabilities occur
(Fig. 1c)
.
As
shown previously in
24
experimental studies on
rate
-
weakening
quartz gouges
,
microstructural evolution and deformation
25
localization
into discre
te
shear bands
is a prerequisite for unstable stick
-
slip behaviour
50
52
. Th
erefore
,
26
in
the
heterogeneous fault
s
,
slip behaviour w
ould
be dictated by the competing processes of fault
27
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
9
stabilization via
deformation
in
weak rate
-
strength
en
i
ng materials
, versus
destabilization caused
by
1
localiz
ation
within the
strong rate
-
weakening patches. When the strong rate
-
weakening patches are
2
large enough
for their internal structur
e
to evolve inde
pe
ndently
,
stick
-
slip instability
may occur.
3
The role of heterogeneity is summarized
in
F
igure 4, where
,
for a given clay
-
qu
artz mixture,
4
heterogeneous faults are weaker and
less stable
relative to their homogeneous equivalents
.
Although it
5
is often invoked that large
-
scale heterogeneities are responsible for the spectrum of slip behaviour
6
observed
on natural faults
17,18
,
the results presented here highlight the
potential of
small
-
scale
7
heterogeneities, which are also abundant in natural fault zones
9,10,22
,
to
exert a significant control on
8
fault zone strength and stability.
There are similarities between the slip behaviour we observe in our
9
small
-
scale heterogeneous experiments and how large
-
scale h
eterogeneities are thought to control the
10
behaviour of natural faults. For example, decreasing the size of the
rate
-
weakening patch makes the
11
response more stable in both our experiments and numerical modelling
53
, as can be intuitively expected
12
and co
nsistent with stability studies of rate
-
and
-
state faults that slip instability can only result from large
13
enough
rate
-
weakening patches
39
. At the same time, small
-
scale fault zo
ne heterogeneity would more
14
readily evolve with shear, and hence may depend on the fault maturity, healing processes, and spatio
-
15
temporal history of fault slip.
16
To summarize, w
e show that
,
by
introducing a simple heterogeneous structure into a fault zone
,
17
the
fault
strength
is
substantially
reduced
and the stability of the experimental fault
is
overall
decrease
d
18
in comparison to compositionally identical b
ut
homogeneously mixed gouges
.
O
ur data, along with the
19
abundance and complexity of
heterogeneity that occurs over many different scales in na
ture
9,10,22
25
,
20
suggest that interactions between heterogeneously distributed materials with different
frictional
21
properties
likely exert
s
an important
control over the mechanical
strength
and
influence
s
whether
22
t
ectonic faults
experience aseismic or earthquake slip
.
The smaller the scale of heterogeneity, the more
23
likely it is to be intractable in modelling
earthquake source
processes and hence ignored
.
Th
ese
24
consideration
s
, together with our findings, necessitate
further laboratory experiments
and modelling to
25
study
the effects and evolution of fault rock heterogeneity within complex fault zones
,
to enable the
26
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
10
quantification
and inclusion of the small
er
-
scale heterogeneity effects into larger
-
scale constitutive laws
1
for modelling fault processes of societal interest,
such as nucleation of natural and induced earthquakes.
2
3
Methods
4
Experimental Procedure
5
The gouge layers are deformed in a direct
-
shear arrangement (
Supplementary
Fig. 1) within a
6
triaxial deformation apparatus
54
. The layers (
~
1
.3
mm initial thickness
prior to pressurization
), prepared
7
in either heterogeneous patches or as a homogeneous quartz
-
clay mixture, are placed between the direct
-
8
shear forcing blocks and soft silicone spacers
are positioned at each end so that displacement can be
9
accommodated without supporting any load (
Supplementary
Fig. 1).
To discourage boundary shear at
10
the edges of the gouge layer, the sliding area (50
×
20 mm) on the forcing blocks contains grooves cut
11
perpendicular to the sliding direction (200 μm deep with 400 μm spacing).
Once the gouge layer is
12
constructed, the direct
-
shear arrangement is surrounded by a low
-
friction polytetrafluoroethylene
13
(PTFE) sleeve (0.25 mm thickness) to minimize jacket frictio
n in the vicinity of the layer, before being
14
placed into a
soft,
3 mm thick
,
PVC jacket
(Nalgene 180 clear tubing)
. The jacketed direct
-
shear
15
arrangement is then placed in between the platens of the sample assembly which is inserted into the
16
pressure vesse
l of the triaxial apparatus. In this geometry, the normal stress
(
)
is applied to the gouge
17
layer by the confining pressure. The pore
-
fluid pressure is introduced to the layer through three porous
18
disks, embedded in each direct
-
shear forcing block, whi
ch are positioned to ensure an even distribution
19
of pore fluid throughout the layer. Deionized water is used as the pore fluid. Both the confining and
20
pore
-
fluid pressures are held constant throughout the experiments by servo
-
controlled pumps on each
21
press
ure system, with a resolution better than 0.01 MPa.
Linear variable differential transformers
22
(LVDTs) are attached to the pistons of the servo
-
control pumps, meaning that the volume of fluid
23
expelled from the sample as it compacts during shearing can be mo
nitored as the pressure is held
24
constant. We therefore use the pore pressure pump as a pore volumometer to track the evolution of layer
25
thickness during our experiments (Supplementary Fig. 2)
;
we assume that sliding area remains constant
26
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
11
and that all volum
etric strain is accommodated by a
change in layer thickness.
The
gouge
layers are
1
sheared by the axial piston of the triaxial apparatus and velocity steps are imposed to calculate the rate
-
2
and
-
state friction parameters. The evolution of shear stress is mon
itored by an internal force gauge
3
within the axial piston, with a measurement resolution of better than 0.05 kN.
4
5
Data
Availability
6
The associated experimental data files for this research can be accessed in National Geoscience Data
7
Center (NGDC) via the following link:
8
https://webapps.bgs.ac
.uk/services/ngdc/accessions/index.html#item164865
9
10
References
11
1.
Lachenbruch, A. H. & Sass, J. H. Heat flow and energetics of the San Andreas Fault Zone.
J.
12
Geophys. Res.
85
, 6185
6222 (1980).
13
2.
Zoback, M. D.
et al.
New Evidence on the state of stress of the San Andreas fault system.
14
Science (80
-
. ).
238
, 1105
1111 (19
87).
15
3.
Copley, A., Avouac, J.
-
P., Hollingsworth, J. & Leprince, S. The 2001 Mw 7.6 Bhuj
16
earthquake, low fault friction, and the crustal support of plate driving forces in India.
J.
17
Geophys. Res.
116
, (2011).
18
4.
Lamb, S. Shear stresses on megathrusts: Impl
ications for mountain building behind subduction
19
zones.
J. Geophys. Res.
111
, (2006).
20
5.
Holdsworth, R. E. Weak Faults
Rotten Cores.
Science (80
-
. ).
303
, 181
182 (2004).
21
6.
Chiaraluce, L., Chiarabba, C., Collettini, C., Piccinini, D. & Cocco, M. Architecture and
22
mechanics of an active low
-
angle normal fault: Alto Tiberina Fault, northern Apennines, Italy.
23
J. Geophys. Res.
112
, (2007).
24
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
12
7.
Byerlee, J. Friction of Rocks.
Pure Ap
pl. Geophys.
116
, 615
626 (1978).
1
8.
Ikari, M. J., Marone, C. & Saffer, D. M. On the relation between fault strength and frictional
2
stability.
Geology
39
, 83
86 (2011).
3
9.
Collettini, C., Niemeijer, A., Viti, C. & Marone, C. Fault zone fabric and fault wea
kness.
4
Nature
462
, 907
911 (2009).
5
10.
Tesei, T., Collettini, C., Barchi, M. R., Carpenter, B. M. & Di Stefano, G. Heterogeneous
6
strength and fault zone complexity of carbonate
-
bearing thrusts with possible implications for
7
seismicity.
Earth Planet. Sci. L
ett.
408
, 307
318 (2014).
8
11.
Niemeijer, A., Marone, C. & Elsworth, D. Fabric induced weakness of tectonic faults.
9
Geophys. Res. Lett.
37
, L03304 (2010).
10
12.
Carpenter, B. M., Marone, C. & Saffer, D. M. Weakness of the San Andreas Fault revealed by
11
samples
from the active fault zone.
Nat. Geosci.
4
, 251
254 (2011).
12
13.
Attanayake, J.
et al.
Rupture characteristics and bedrock structural control of the 2016 Mw 6.0
13
intraplate earthquake in the Petermann Ranges, Australia.
Bull. Seismol. Soc. Am.
110
, 1037
14
104
5 (2020).
15
14.
Di Toro, G.
et al.
Fault lubrication during earthquakes.
Nature
471
, 494
498 (2011).
16
15.
Rice, J. R. Fault stress states, pore pressure distributions, and the weakness of the San Andreas
17
Fault. in
Fault Mechanics and Transport Properties of R
ocks
(eds. Evans, B. & Wong, T.) vol.
18
51 475
503 (Academic Press, 1992).
19
16.
Faulkner, D. R. & Rutter, E. H. Can the maintenance of overpressured fluids in large strike
-
20
slip fault zones explain their apparent weakness?
Geology
29
, 503
506 (2001).
21
17.
Avouac, J.
-
P. From geodetic imaging of seismic and aseismic fault slip to dynamic modeling
22
of the seismic cycle.
Annu. Rev. Earth Planet. Sci.
43
, 233
271 (2015).
23
18.
Thomas, M. Y., Avouac, J.
-
P., Champenois, J., Lee, J.
-
C. & Kuo, L.
-
C. Spatiotemporal
24
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
13
evol
ution of seismic and aseismic slip on the Longitudinal Valley Fault, Taiwan.
J. Geophys.
1
Res. Solid Earth
119
, 5114
5139 (2014).
2
19.
Bürgmann, R. The geophysics, geology and mechanics of slow fault slip.
Earth Planet. Sci.
3
Lett.
495
, 112
134 (2018).
4
20.
Co
llettini, C., Tesei, T., Scuderi, M. M., Carpenter, B. M. & Viti, C. Beyond Byerlee friction,
5
weak faults and implications for slip behavior.
Earth Planet. Sci. Lett.
519
, 245
263 (2019).
6
21.
Barnes, P. M.
et al.
Slow slip source characterized by lithologi
cal and geometric heterogeneity.
7
Sci. Adv.
6
, (2020).
8
22.
Fagereng, Å. Frequency‐size distribution of competent lenses in a block‐in‐matrix mélange:
9
Imposed length scales of brittle deformation?
J. Geophys. Res.
116
, (2011).
10
23.
Faulkner, D. R., Lewis, A.
C. & Rutter, E. H. On the internal structure and mechanics of large
11
strike
-
slip fault zones: field observations of the Carboneras fault in southeastern Spain.
12
Tectonophysics
367
, 235
251 (2003).
13
24.
Faulkner, D. R.
et al.
A review of recent developments co
ncerning the structure, mechanics
14
and fluid flow properties of fault zones.
J. Struct. Geol.
32
, 1557
1575 (2010).
15
25.
Rutter, E. H., Faulkner, D. R. & Burgess, R. Structure and geological history of the Carboneras
16
Fault Zone, SE Spain: Part of a stretchin
g transform fault system.
J. Struct. Geol.
45
, 68
86
17
(2012).
18
26.
Bilek, S. L. & Lay, T. Tsunami earthquakes possibly widespread manifestations of frictional
19
conditional stability.
Geophys. Res. Lett.
29
, 1673 (2002).
20
27.
Kirkpatrick, J. D.
et al.
Subductio
n megathrust heterogeneity characterized from 3D seismic
21
data.
Nat. Geosci.
13
, 369
374 (2020).
22
28.
Crawford, B. R., Faulkner, D. R. & Rutter, E. H. Strength, porosity, and permeability
23
development during hydrostatic and shear loading of synthetic quartz
-
clay fault gouge.
J.
24
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
14
Geophys. Res.
113
, (2008).
1
29.
Tembe, S., Lockner, D. A. & Wong, T.
-
F. Effect of
clay content and mineralogy on frictional
2
sliding behavior of simulated gouges: Binary and ternary mixtures of quartz, illite, and
3
montmorillonite.
J. Geophys. Res.
115
, (2010).
4
30.
Takahashi, M., Mizoguchi, K., Kitamura, K. & Masuda, K. Effects of clay c
ontent on the
5
frictional strength and fluid transport property of faults.
J. Geophys. Res.
112
, (2007).
6
31.
Kenigsberg, A. R., Rivière, J., Marone, C. & Saffer, D. M. Evolution of elastic and mechanical
7
properties during fault shear: The roles of clay cont
ent, fabric development, and porosity.
J.
8
Geophys. Res. Solid Earth
125
, (2020).
9
32.
Skarbek, R. M. & Savage, H. M. RSFit3000 : A MATLAB GUI
-
based program for
10
determining rate and state frictional parameters from experimental data.
Geosphere
15
, 1665
11
1676
(2019).
12
33.
Faulkner, D. R., Sanchez
-
Roa, C., Boulton, C. & den Hartog, S. A. M. Pore fluid pressure
13
development in compacting fault gouge in theory, experiments, and nature.
J. Geophys. Res.
14
Solid Earth
123
, 226
241 (2018).
15
34.
Rutter, E. H.
et al.
Reduct
ion of friction on geological faults by weak
-
phase smearing.
J.
16
Struct. Geol.
51
, 52
60 (2013).
17
35.
Knuth, M. W., Tobin, H. J. & Marone, C. Evolution of ultrasonic velocity and dynamic elastic
18
moduli with shear strain in granular layers.
Granul. Matter
15
,
499
515 (2013).
19
36.
Dieterich, J. H. Modeling of rock friction 1. Experimental results and constitutive equations.
J.
20
Geophys. Res.
84
, 2161
2168 (1979).
21
37.
Dieterich, J. H. Constitutive properties of faults with simulated gouge. in
Mechanical Behavior
22
o
f Crustal Rocks
(eds. Carter, N., Friedman, M., Logan, J. & Stearns, D.) 103
120 (1981).
23
doi:https://doi.org/10.1029/GM024p0103.
24
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
15
38.
Ruina, A. Slip instability and state variable friction laws.
J. Geophys. Res.
88
, 10359
10370
1
(1983).
2
39.
Rice, J. R. & Rui
na, A. L. Stability of steady frictional slipping.
J. Appl. Mech.
50
, 343
349
3
(1983).
4
40.
Scholz, C. H. Earthquakes and friction laws.
Nature
391
, 37
42 (1998).
5
41.
Marone, C. Laboratory
-
derived friction laws and their application to seismic faulting.
Annu.
6
Rev. Earth Planet. Sci.
26
, 643
696 (1998).
7
42.
Buijze, L., Guo, Y., Niemeijer, A. R., Ma, S. & Spiers, C. J. Effects of heterogeneous gouge
8
segments on the slip b
ehavior of experimental faults at dm scale.
Earth Planet. Sci. Lett.
554
,
9
116652 (2021).
10
43.
Fagereng, Å. & Sibson, R. H. Mélange rheology and seismic style.
Geology
38
, 751
754
11
(2010).
12
44.
Smeraglia, L.
et al.
Ultra
-
thin clay layers facilitate seismic sli
p in carbonate faults.
Sci. Rep.
7
,
13
(2017).
14
45.
Chester, F. M. & Chester, J. S. Ultracataclasite structure and friction processes of the
15
Punchbowl fault, San Andreas system, California.
Tectonophysics
295
, 199
221 (1998).
16
46.
Skarbek, R. M., Rempel, A. W.
& Schmidt, D. A. Geologic heterogeneity can produce
17
aseismic slip transients.
Geophys. Res. Lett.
39
, (2012).
18
47.
Noda, H. & Lapusta, N. Stable creeping fault segments can become destructive as a result of
19
dynamic weakening.
Nature
493
, 518
521 (2013).
20
48.
Bayart, E., Svetlizky, I. & Fineberg, J. Rupture dynamics of heterogeneous frictional
21
interfaces.
J. Geophys. Res. Solid Earth
123
, 3828
3848 (2018).
22
49.
Bahr, D. B. & Rundle, J. B. Stick
-
slip statistical mechanics at the bed of a glacier.
Geophys.
23
Res. L
ett.
23
, 2073
2076 (1996).
24
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
16
50.
Scuderi, M. M., Collettini, C., Viti, C., Tinti, E. & Marone, C. Evolution of shear fabric in
1
granular fault gouge from stable sliding to stick slip and implications for fault slip mode.
2
Geology
45
, 731
734 (2017).
3
51.
Bedfor
d, J. D. & Faulkner, D. R. The role of grain size and effective normal stress on
4
localization and the frictional stability of simulated quartz gouge.
Geophys. Res. Lett.
48
,
5
e2020GL092023 (2021).
6
52.
Gu, Y. & Wong, T.
-
F. Development of shear localization i
n simulated quartz gouge: effect of
7
cumulative slip and gouge particle size.
Pure Appl. Geophys.
143
, 387
423 (1994).
8
53.
Chen, T. & Lapusta, N. Scaling of small repeating earthquakes explained by interaction of
9
seismic and aseismic slip in a rate and state fault model.
J. Geophys. Res.
114
, B01311 (2009).
10
54.
Faulkner, D. R. & Armitage, P. J. The effect of tectonic environme
nt on permeability
11
development around faults and in the brittle crust.
Earth Planet. Sci. Lett.
375
, 71
77 (2013).
12
13
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
17
Acknowledgements
1
Gary Coughlan is thanked for assistance in developing and maintaining the experimental apparatus.
2
We are
grateful to Elisabetta Mariani for help with and maintenance of the SEM facilities. This work
3
is supported by Natural Environment Research Council grant NE/P002943/1.
4
5
Author contributions
6
J.D.B and D.R.F developed the main ideas. J.D.B performed the exper
iments, ran microstructural
7
analyses and produced the initial manuscript. All authors contributed to interpreting the results and
8
editing the manuscript.
9
10
Competing interests
11
The authors declare no competing interests.
12
13
Materials and correspondence
14
Correspondence and material requests should be addressed to J.D.B.
15
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
18
1
Figure 1|
Mechanical behaviour of
laterally
heterogeneous vs. homogeneously
mixed clay
-
quartz
2
fault gouge layers
. a,
Schematic diagram
of a
typical
natural fault
zone
showing how lenses
of different
3
materials trapped within the fault core produces a heterogeneous structure.
b,
Simplified diagrams of
4
the experimental setup for the heterogenous fault experiments, where quartz and clay gouges are
5
separated into adjacent patches, and homogenous fault experiments where the two gouges are
6
homogenously mixed together.
c
,
Evolution of
the friction coefficient (
μ
)
with displacement for the
7
heterogeneous experimental faults and,
d,
the homogeneous experimental faults.
Heterogeneous faults
8
show ubiquitous
post
-
yield
weakening
with increasing displacement
,
in
contrast
to homogeneous faults
9
where
μ
remains relatively constant once the layer has yielded after approximately 1 mm of slip. For
10
pure quartz and heterogeneous faults
where the quartz patch comprises
≥80% of the total fault area
,
11
stick
-
slip instabilities occur, triggered by up
-
steps i
n the sliding velocity.
12
13
Th
is
paper is now
published at:
https://doi.org/10.1038/s41467
-
022
-
27998
-
2
19
1
Figure 2| Microstructural evolution and
potential causes of
weakening
in
the heterogeneous fault
2
gouge layers. a,
Backscatter electron image of the interface between a clay
-
quartz patch recovered at
3
the end of an experiment. The clay phase becomes smeared along a boundary Y
-
shear plane that
4
propagates into the quartz patch.
Since it is
difficult
to
keep the gouge laye
r intact upon removal from
5
the direct shear assembly at the end of the experiment, the full extent of the localized shear band was
6
not recovered.
b,
Schematic diagram showing
the evolution of the fault gouge layers with
progressive
7
smearing of the clay pha
se along localized Y
-
boundary shears
(
red box shows
the location of the
8
micrograph in
a
)
.
c,
Observed weakening versus predicted weakening
due to clay smearing
for
9
heterogeneous layers comprised of 30 and
7
0% clay fractions. The predicted weakening is calc
ulated
10
using the arithmetic mean of the
friction coefficients of the
endmember quartz and clay gouges and by
11
assuming that the length of the clay patches increases by the amount of displacement on the fault as
12
clay is smeared along localized Y
-
shear planes
. The observed weakening is considerably greater than
13
the predicted weakening. The labels (I), (II) and (III) correspond to the structural evolution in
b
.
d
, The
14
potential weakening effect from differential compaction between the clay and quartz gouge patc
hes.
15
This effect is dependent on the
b
ulk (
) and shear (
) moduli of the gouge
,
which are poorly constrain
ed
16
(see Supplementary Information for full discussion).
The
differential compaction could account for a
17
large component of the weakening in the heterogeneous fault experiments.
18
19