Journal of Structural Geology 32 (2010) 2042e2071

Contents lists available at ScienceDirect

Journal of Structural Geology

journal homepage: [Link]/locate/jsg

Review Article

Information from folds: A review

Peter J. Hudleston a, *, Susan H. Treagus b
a
Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, USA
b
School of Earth, Atmospheric and Environmental Sciences, University of Manchester, Manchester M13 9PL, UK, England

a r t i c l e i n f o a b s t r a c t

Article history: Folds are spectacular geological structures that are seen in layered rock on many different scales. To mark
Received 13 January 2010 30 years of the Journal of Structural Geology, we review the information that can be gained from studies
Received in revised form of folds in theory, experiment and nature. We first review theoretical considerations and modeling, from
24 July 2010
classical approaches to current developments. The subject is dominated by single-layer fold theory, with
Accepted 22 August 2010
Available online 8 September 2010
the assumption of perfect layer-parallel shortening, but we also review multilayer fold theory and
modeling, and folding of layers that are oblique to principal stresses and strains. This work demonstrates
that viscosity ratio, degree of non-linearity of the flow law, anisotropy, and the thickness and spacing
Keywords:
Folds
distribution of layers of different competence are all important in determining the nature and strength of
Folding the folding instability. Theory and modeling provide the basis for obtaining rheological information from
Deformation history natural folds, through analysis of wavelength/thickness ratios of single layer folds, and fold shapes. They
also provide a basis for estimating the bulk strain from folded layers. Information about folding mech-
anisms can be obtained by analysis of cleavage and fabric patterns in folded rocks, and the history of
deformation can be revealed by understanding how asymmetry can develop in folds, by how folds
develop in shear zones, and how folds develop in more complex three-dimensional deformations.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction terminology and Van Hise (1894) on geometry. Much of the work
on folds in the first half of the 20th century was concerned with
Folds are spectacular structures in deformed rocks, affecting developing geometrical methods for representing folds. Details can
single or multiple layers on all scales, and on a small-scale be found in the textbooks of Leith (1923), Nevin (1931), Hills (1963)
commonly seen affecting veins, schistosities and foliations (Figs. 1 and de Sitter (1964). A review of work on the mechanics of folding
and 2). They have played an important part, historically, in under- through the mid 1970s can be found in Johnson (1977: Chapter 1).
standing episodes of deformation in orogenic belts. To mark 30 In this review, we concentrate on the information that folded
years of the Journal of Structural Geology, we combine forces and rocks and their analysis can provide: information on rheology,
indulge our separate love of folds to review the information that strain and deformation history, locally or regionally. Much of this
can be gained from studies of folds in theory, experiment and information stems from developments in fold theory and modeling
nature. over the course of the last 50 years, beginning with the work of Biot
It was probably Hall (1815) who first used the word folds in (1961, 1965a, 1965b), who developed theories for single and
connection with rock structures. He was describing the models he multilayer folding in viscoelastic and viscous media, with applica-
had made from layered cloths confined between boards and tions to rocks, Currie et al. (1962), who developed models of elastic
laterally compressed, to simulate folded rocks he had observed on folding and structural lithic units that were related to folds in the
the Berwickshire coast of Britain. He wrote: “The consequence was Appalachians, and Ramberg (1959, 1963, 1964, 1970), who made
. the strata were constrained to assume folds, bent up and down, significant contributions to the modern understanding of folding
which very much resemble the convoluted beds . exhibited in the mechanisms, based on theory and model experiments. These
crags of Fast Castle”. Among the early studies of folds are the studies together form the foundations of modern fold theory and
outstanding contributions of Willis (1891) on mechanics and modeling, expanded below.
In their papers on buckling, Biot (1961, 1965a) and Ramberg
(1964, 1970, 1981) extended their analysis to include the influ-
* Corresponding author.
ence of gravity. In this paper we do not consider gravitational
E-mail addresses: hudle001@[Link] (P.J. Hudleston), [Link]@ forces, which may become important for large folds, and are
[Link] (S.H. Treagus). certainly so for folds that affect the earth’s surface. In the last

0191-8141/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/[Link].2010.08.011
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2043

Fig. 1. Examples of small-scale buckle folds. (a) Quartz veins in slate, Trondheim, Norway; scale bar 20 cm. (b) Quartz veins in schist, Cap de Creus, Spain; coin 2 cm. (c) Mylonitized
pegmatite vein in mylonites, Cap de Creus, Spain, showing wavelength decreasing with thickness; scale bar 10 cm. (d). Single-layer buckling in thin white pegmatitic veins modified
by multilayer effects of banding in gneisses of the Maggia nappe, Ticino, Switzerland; coin 2.3 cm.

Fig. 2. (a) Multilayer folds in Moine schists, with wavelengths determined by competent white quartzo-feldspathic veins of different thickness (v) that have buckled largely
independently; Loch Monar, Scotland; scale bar 20 cm. (b) Multilayer folds in anisotropic gneiss, Maggia nappe, Ticino, Switzerland; hand lens 5 cm. (c) Multilayer folding in the
New Harbour Formation psammitic schists, Silver Bay, Anglesey, UK; lens cap 5 cm. (d) Multilayer buckle folds of chevron style and variable asymmetry in siltstone and slates,
Boscastle, Cornwall, England; scale bar 20 cm. The overall fold style is similar (Class 1C, Ramsay, 1967, p. 367), but the stiff siltstone layers have parallel shapes (Class 1B).
2044 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

decade or so, work on large-scale buckle folding has focused on the (3)) depends on the ratio of the viscous to elastic dominant
whole lithosphere (e.g. Burg and Podladchikov, 1999; Cloetingh wavelengths. If the ratio, R ¼ ldv/lde < 1, ld is given approximately
et al., 2002), at which scale gravitational forces are of great by Eq. (1) and if R ¼ ldv/lde > 1, ld is given approximately by Eq. (3).
importance. A buckling instability is in fact one of a family of dynamic
instabilities that result from either compression or extension of an
isolated layer that is either more or less viscous than its matrix, as
2. Theoretical considerations and modeling
shown by Smith (1975, 1977). Folding is by far the strongest of these
four instabilities and the only one considered here. The other
2.1. Single-layer fold theory
important instability is that of pinch and swell or boudinage that
results from extension of a stiff layer in a less competent matrix,
Single layer fold theory concerns the buckling of isolated layers
and this instability is only significant if the material behavior is
subjected to layer-parallel compression, developed for the case of
non-linear, because for Newtonian materials the dynamic growth is
a stiff or competent viscous layer in a less stiff or less competent
counteracted by the kinematic decay associated with the base flow.
matrix (Biot, 1961; Biot et al., 1961; Ramberg, 1961, 1963; Chapple,
However, if the base flow involves extension along the axes of the
1968; Fletcher, 1974, 1977; Smith, 1975, 1977, 1979; Johnson and
boudins, it may be possible to produce boudinage in Newtonian
Fletcher, 1994) and for the corresponding cases of elastic and
flow (James and Watkinson, 1994).
viscoelastic layers and matrix (Biot, 1961, 1965a; Currie et al., 1962;
Thin-plate analysis takes into account only normal stresses in the
Johnson, 1977; Mühlhaus et al., 1994, 1998; Hunt et al., 1996, 1997;
competent layer, and the stresses associated with folding consist of
Schmalholz and Podladchikov, 1999, 2000; Jeng and Huang, 2008).
deviatoric tension in the outer arcs and deviatoric compression in
Classical theory predicts that if the layer is given small sinusoidal
the inner arcs (Fig. 3). Layer-parallel shear stress is ignored. Thick
perturbations of different wavelengths, one such perturbation will
plate analysis does not make this restrictive assumption about
amplify at a greater rate than all others. The wavelength of this
stresses (Johnson and Fletcher, 1994, pp. 196e207), and thus
perturbation is termed the dominant wavelength, ld. For Newtonian
provides more accurate predictions of the dominant wavelength at
viscous layer and matrix in plane strain, with maximum shortening
low viscosity contrast and for low-amplitude fold growth. Third
parallel to the layer and ignoring gravity and inertial effects, ld
order analysis of the problem (Johnson and Fletcher, 1994, pp.
depends only on the ratio of viscosities of layer to matrix. The thin-
224e236) accounts for the deviation of the layer shape of the
plate (see below) approximation for ld is:
growing fold from sinusoidal. Thick plate analysis does not lead

1=3 to an explicit expression for the dominant wavelength. In
ld mL
¼ 2p ; (1) viscous materials, whether thin-plate or thick-plate analysis is
h 6mM used, the layer will shorten and thicken under the influence of the
where h is layer thickness and mL and mM the viscosities of layer and compressive normal stress (Fig. 4), and treating this analytically
matrix (Biot, 1961; Ramberg, 1961). This approximation holds for involves considering the stresses in the layer to consist of two parts,
both welded and free-slip contacts between layer and matrix. It is a uniform stress that results in shortening of the layer with time
good for mL/mM  100, but becomes increasingly inaccurate as mL/mM and a perturbation stress field that is the source of the buckling
is decreased and the assumptions of the thin plate formulation instability (see Fig. 3 for the thin-plate case). At high viscosity
become untenable. (Note in this regard that Eq. (1) gives a domi- contrasts the amount of shortening that will occur before the folds
nant wavelength for the case when mL/mM ¼ 1, which is not physi- grow to finite amplitude is small, but at low viscosity contrasts
cally meaningful). significant layer-parallel shortening will occur as the folds amplify.
Under suitable conditions, including sufficiently large strain
rates, the elastic properties of rocks may influence the folding
instability. The thin-plate expression for the dominant wavelength
of an elastic layer in an elastic matrix is identical in form to Eq. (1)
(e.g., Currie et al., 1962; Jeng and Huang, 2008).

1=3
ld EL
¼ 2p ; (2)
h 6EM
where EL and EM are the elastic moduli of layer and matrix. Elastic
behavior by itself is obviously inappropriate for rocks, in which
folds represent permanent inelastic deformation. If an elastic layer
is embedded in a viscous matrix, the dominant wavelength is
dependent on applied load (or alternatively rate of deformation). It
is given by:
!1=2
ld EL
¼ p   ; (3)
h P 1  y2L

where nL is Poisson’s ratio and P is the layer-parallel stress in the

stiff layer (Biot, 1961; Turcotte and Schubert, 1982). Note that the
dominant wavelength in this case is independent of the viscosity of
the matrix. If the layer is viscoelastic (Maxwell rheology, equivalent
to a spring and a dashpot in series) and the matrix viscous,
Schmalholz and Podladchikov (1999) showed that whether the Fig. 3. In thin-plate analysis, the normal stress, sx, in the layer can be considered the
folding is controlled largely by its viscous properties (with ld given sum of a uniform (membrane) stress, sx , and fiber stresses, s ~x , related to the pertur-
by Eq. (1)) or largely by its elastic properties (with ld given by Eq. bation of the layer (after Schmalholz and Podladchikov, 2000).
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2045

expression for the dominant wavelength corresponding to the thin-

plate expression, Eq. (1), is (Fletcher, 1974):
!1=3
ld m0L n1=2
¼ 2p M
; (6)
a h 6m0M nL

where nL and nM are the power-law exponents and m0 L and m0 M the

viscosities under the base rate of flow (in response to the stress sx in
Fig. 3) for the layer and matrix respectively. Note that for power-law
fluids it is not meaningful to talk about a single viscosity contrast
between two rock types because viscosity varies as function of
stress or strain rate, and thus viscosity ratio will vary as stress or
b strain-rate vary (Treagus, 1993).
Another important rock property that affects the folding insta-
bility is anisotropy. As pointed out by Fletcher (1974), power-law
rheology and anisotropy of the competent layer, in which the
viscosity of the layer in shear parallel to the layer is less than the
viscosity in shortening or extension, have opposite effects on
the secondary flow associated with folding and thus on the buck-
ling instability. The stress-induced anisotropy of the secondary flow
associated with power-law rheology involves the layer becoming
c weaker in compression than in shear, whereas the intrinsic
anisotropy of layered rocks involves the layer being weaker in shear
Fig. 4. Definition of terms used to described periodic folds at various stages of parallel to the layer than in compression. Anisotropy of this type in
development. (a) Initial state; (b) after fold nucleation and growth and towards the end
of wavelength selection; (c) after growth to large amplitude. If L z lp at stage (b), L in
the competent layer lessens the buckling instability, whereas
(c) is a rough measure of lp at the strain shown in (b) (see text). anisotropy in the matrix increases it, compared to the isotropic case
and if normal viscosities of layer and matrix are kept fixed (Kocher
et al., 2006).
Ramberg (1964) emphasized the implications of this for single In general, for viscous materials the rate of growth of a sinu-
layers with different values of mL/mM deformed under the same soidal perturbation can be expressed by a function of the following
conditions of overall shortening e the layers will appear to have form (Fletcher, 1974):
undergone different degrees of folding. At very low viscosity
dA
contrasts the kinematic amplification, due only to the geometric ¼ ½1 þ qðkÞAðkÞ; (7)
ds
change associated with homogeneous strain, masks the dynamic
amplification due to the buckling instability. where A is amplitude, q(k) is a growth factor, k ¼ 2ph=l, and s is
Sherwin and Chapple (1968) modified thin plate theory to take dimensionless time (s ¼ e_ x t, where e_ x is the base strain rate
into account the effect of layer-parallel shortening that accom- corresponding to the stress sx , t is time and e2s ¼ T). This implies
panies low-amplitude fold growth. In this case, the wavelength of an exponential increase in fold amplitude with time or with
the folds with the greatest cumulative amplification is a function of shortening strain. Using e2s ¼ T and changing and separating
the layer-parallel shortening. Johnson and Fletcher call this the variables the expression for amplification becomes:
preferred wavelength, lp. The approximate (thin-plate) expression
for lp is: ZT
A 1
ln ¼ ½1 þ qðkÞ dT; (8)

 Ao 2T
lp mL T þ 1 1=3 1
¼ 2p ; (4)
hðTÞ 6mM 2T 2 where Ao is initial amplitude. The value of q(k) depends on the
rheological properties and implies mechanical instability if
where T ¼ S1/S2 ¼ S21, and S1 and S2 are the principal stretches greater than zero. A value of q ¼ 0 describes purely kinematic
perpendicular and parallel to the layer respectively. From Biot amplification. For a given set of rheological parameters, q(k)
(1965a, p. 427), and Johnson and Pfaff (1989) it can be shown that: attains a maximum value for the dominant wavelength. Exact
first order expressions for q(k) have been derived for linear
lp ld
¼ S2 : (5) rheology under conditions of bonded-contact and free-slip layer
h ho
interfaces by Johnson and Fletcher (1994, pp. 196e206) and for
The analyses discussed above assume a form of linear viscous or bonded-contact power-law rheology (Fletcher, 1974; Johnson and
viscoelastic rheology. However, experiments on common rock- Fletcher, 1994, p. 383). Results for linear rheology, bonded
forming minerals and rocks, and fabric analyses in naturally- contacts and three values of mL/mM are shown in Fig. 5a, and
deformed rock, suggest that under certain conditions of ductile values of the dominant wavelength/thickness, ld/h, as a function
deformation, rocks might be expected to follow non-linear flow of viscosity contrast are shown in Fig. 5b, where the differences
laws (e.g. Carter and Tsenn, 1987; Kirby and Kronenberg, 1987; between thin-plate and exact solutions with either bonded or
Rutter, 1993; Kohlstedt et al., 1995; Hirth et al., 2001). Fletcher free-slip contacts can be compared. Thin-plate theory over-
(1974) and Smith (1975, 1977) independently developed the estimates the growth rate and underestimates the value of ld/h
theory for folding of layers with non-linear rheology, in the case of for a given viscosity contrast.
Fletcher specifically power-law, for which the single component The variations of q(k) with l/h for a viscosity ratio of 50 and
stress e strain rate relationship is of the form e_ ¼ Bsn , where n is for both linear and power-law behavior of the layer and isotropic
the power-law exponent and B a constant. An approximate and anisotropic behavior of the matrix are shown in Fig. 6. The
2046 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

a b

Fig. 5. (a) Growth factor, q(k) (Johnson and Fletcher, 1994, Eq. 5.2.12b) as a function of l/h for linear viscous layer and matrix (bonded contacts) and three different viscosity
contrasts. (b) Dominant wavelength/thickness, ld/h, as a function of viscosity ratio for an isolated single layer, for the thin-plate (Biot) approximation and for exact thick-plate first
order solutions with either bonded or free-slip contacts and linear viscosity (Eq. (1), and from Johnson and Fletcher, 1994, Eqs. 5.2.16b, 5.2.17).

growth rate curves in this figure were obtained from Fletcher obviously depends on the power-law exponent and the degree of
(1974, Eq. 8): anisotropy. An analytical solution for an anisotropic layer in an
isotropic matrix has been derived by Fletcher (personal commu-
 n   pffiffiffiffiffiffiffiffiffiffiffiffiffiffih
qðkÞ ¼ 2nL 1  R0  1  Q 2 þ nL  1 ð1 þ Q Þ2 eaL k nication), and this and numerical simulations (Kocher et al., 2006,
i. o1 Fig. 2c), show that the growth rate for an anisotropic layer is sup-
 ð1  Q Þ2 eaL k 2sinðbL kÞ ; (9) pressed compared to the linear case, for fixed mnM =mnL .
qffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffi 00 pffiffiffiffiffiffiffiffiffiffiffi Using the appropriate expression for q(k), Eq. (8) can be inte-
where R0 ¼ m0n 0n
M =mL , R ¼
00 m0n m0s m0n
M M= L , Q ¼ nL R , aL ¼ 1=nL ; grated numerically over the path k(s) ¼ koe2s ¼ koT to find the
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
and bL ¼ 1  1=nL . Fletcher (1974) pointed out, following Biot cumulative amplification, A/Ao, of folds with initial wave number,
(1965a, p. 211), that for single-layer folding, an anisotropic ko. By creating spectra of A/Ao as a function of k the value
viscous matrix behaves like an isotropic viscous half-space with kp ¼ 2ph=lp that shows the maximum cumulative amplification
pffiffiffiffiffiffiffiffiffiffi
effective viscosity m ¼ mn ms. This allows Eq. (9) to be used to can be found. An example of amplification spectra for increasing
incorporate the effect of an anisotropic matrix. It is clear from Fig. 6 amounts of shortening is shown in Fig. 7 for Newtonian layer and
that power-law behavior of the layer both enhances the growth rate matrix. An exact expression for kp or lp/h was derived by Sherwin
and decreases the dominant wavelength compared to linear and Chapple (1968) for the thin-plate linear viscous case (Eq. (4)).
behavior, whereas anisotropy of the matrix increases both the
growth rate and the dominant wavelength. A combination of 2.2. Finite amplitude single-layer fold theory
power-law layer and anisotropic matrix greatly enhances growth
rates and modestly affects the dominant wavelength. The degree to First-order buckling theory predicts an exponential increase in
which the growth rate and dominant wavelength are affected amplitude with time or shortening (Eqs. (7) or (8)) as folding
initiates and is strictly applicable only for infinitesimal amplitudes.
Extending theoretical analysis beyond first order (Fletcher, 1979;
Johnson and Pfaff, 1989; Johnson and Fletcher, 1994, pp.
224e236), shows that higher waveforms spontaneously develop
that serve to modify fold shape and limit the growth rate of the first
waveform. Cruikshank and Johnson (1993) developed a method of
matching boundary conditions along irregular interfaces that
allows fold development to be predicted accurately to high
amplitudes. Schmalholz and Podladchikov (2000) developed
a simple modification of the thin-plate theory for single layers that
effectively tracks fold amplitude, without specifying the exact fold
shape. This takes into account the fact that for fold growth to
continue exponentially, according to the linear theory, a stretching
of the competent layer would have to occur, and it is resistance to
this stretching that slows fold growth. They assumed that the layer-
parallel stress (membrane stress, s in Fig. 3) in the competent layer
remains constant around the layer as the fold grows to finite
amplitude but decreases with time in response to the resistance of
the layer to stretching. The membrane stress is related to layer-
Fig. 6. Effect of power-law rheology of layer and anisotropy of matrix on dynamic parallel (arclength) strain rate, which is therefore also constant
growth rates and dominant wavelength (given by l/h for the peak values of these around the layer. Both membrane stress and strain rate will
curves) of single layer buckle folds. nL is the power-law exponent of the layer and
dM ¼ m0 nM/m0 sM (with m0 nM and/m0 sM being the normal and shear viscosities of the matrix) is
decrease as the fold grows, and analytical expressions for these
the anisotropy factor of the matrix (see Kocher et al., 2006). In all cases the ratio of were obtained by Schmalholz and Podladchikov (2000), expressing
normal viscosity of the layer to the normal viscosity of the matrix, m0 nL /m0 nM ¼ 50. membrane stress as a function of initial stress and arc strain rate as
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2047

Fig. 9. Schmalholz and Podladchikov (2000) proposed the term

crossover amplitude for the amplitude at which this transition
occurs. They also proposed the term nucleation amplitude, which is
the amplitude at which the dynamic rate of fold growth is equal to
the kinematic rate. Both nucleation amplitude and crossover
amplitude have precise definitions, and depend only on the growth
rate factor, q(k), (Schmalholz and Podladchikov, 2000, Eq. (16);
Schmalholz and Podladchikov, 2001, Eq. (8)).

2.3. Single-layer fold modeling: analog and numerical

Considerable experimental work has been done to investigate

the nature of folding instabilities in single stiff layers embedded in a
less stiff matrix and subjected to layer-parallel shortening, using a
variety of analog materials for layer and matrix. The existence
of a characteristic or quasi-characteristic wavelength in physical
model experiments, and good correspondence with theoretical
predictions, have been found experimentally for elastic layers in an
elastic matrix (Biot et al., 1961; Currie et al., 1962), elastic layers in
viscous matrix (Biot et al., 1961) and viscous layers in viscous
matrix, both for Newtonian (Biot et al., 1961; Hudleston, 1973a) and
non-Newtonian (Neurath and Smith, 1982; Mancktelow and
Fig. 7. Amplification spectra for mL/mM ¼ 50 (Newtonian, isotropic) after increasing Abbassi, 1992) cases.
amounts of shortening strain, T ¼ S1/S2, during low-amplitude fold growth. The black Since digital computers became widely available in the 1960s,
dot shows the dominant wavelength/thickness, ld/h, and the dash-dotted line tracks numerical methods have been applied with considerable success to
the change in preferred wavelength/thickness, lp/h, with shortening. The dashed lines study the development of folds to high amplitude. Chapple (1968)
track wavelengths that have been amplified to half the maximum value. They show the
used a finite difference approach for this purpose, predicting fold
increased selectivity of amplification with shortening.
shape in a thin linearly viscous layer of constant length embedded
in a viscous matrix. Dieterich (1969) and Dieterich and Carter
a function of initial or background strain rate. A comparison of first
(1969) used finite element models (FEM) to analyze fold shape,
order (small amplitude) theoretical predictions of fold growth with
strain distribution and stress history in folds grown to high
the predictions of Schmalholz and Podladchikov’s (2000) finite
amplitudes in viscous materials, free from the assumptions of thin-
amplitude theory is made in Fig. 8. The finite amplitude theory
plate theory. Subsequent numerical studies of single-layer folds
agrees very well with numerical simulations of buckling (see
have included the effects of non-linear rheology (Parrish, 1973; Lan
Schmalholz and Podladchikov, 2000, Fig. 6), which do not depend
and Hudleston, 1991), anisotropy (Lan and Hudleston, 1996; Kocher
on thin-plate assumptions about stress in the competent layer. The
et al., 2006), and viscoelastic behavior (Zhang et al., 1996;
amount of strain accumulated in the stage during which the
Mancktelow, 1999; Schmalholz et al., 2001).
exponential curves track fold growth depends on the amplitude of
Numerical models provide a good means of examining the effects
the initial perturbation in the layer e the smaller the initial
of varying rheological properties and strain rate in a controlled way
perturbation, the greater the amount of strain. In general it is quite
in viscous and viscoelastic media and also provide an excellent way
limited.
of testing the dependence of finite fold shape on the initial layer
The departure of the finite-amplitude growth curve from the
configuration, and in particular on the form of the initial layer
infinitesimal-amplitude growth curve is reflected in a transition
irregularities. It is worth noting that essentially identical results
from layer shortening accompanying exponential growth to fold
growth with little change in layer (arc) length. This is shown in

Fig. 9. Change in relative arclength as a function of strain during folding of a single

Fig. 8. Growth of fold amplitude/wavelength, A/l, for a single layer with mL/mM ¼ 50 Newtonian layer in Newtonian matrix with an initial sinusoidal (dominant) waveform
and with initial sinusoidal perturbation of the dominant wavelength (ld/h ¼ 12.75) and initial amplitude of 0.02 of layer thickness, according to the finite amplitude
and initial amplitude equal to 0.02 of layer thickness, according to thin-plate theory theory of Schmalholz and Podladchikov (2000), for three values of viscosity ratio, mL/mM.
(Biot, 1961), thick-plate theory (Fletcher, 1974) and finite amplitude theory of The line for “no folding” represents uniform shortening of the layer without folding. The
Schmalholz and Podladchikov (2000). The predictions of the finite amplitude theory black dots indicate the crossover strains, given by stretch Sc (Schmalholz and
are closely matched by 2D numerical simulations (Schmalholz and Podladchikov, Podladchikov, 2000), that approximately limit the stage of folding described by the
2000). linear theory.
2048 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

have been obtained for a wide range of conditions using different Schmalholz and Podladchikov (1999) confirmed with numerical
numerical techniques. One apparent conflict between results using models their theoretical prediction that the wavelength of the folds
two different techniques studying buckling in viscoelastic media developed in a viscoelastic layer in a viscous matrix depends on the
(see Zhang et al., 1996, who used a finite difference code; and ratio, R ¼ ldv/lde, of viscous dominant wavelength (Eq. (1)) to elastic
Mancktelow, 1999, who used a finite element code) was due to the dominant wavelength (Eq. (3)): viscous control when R < 1 and
fact that different strain rates had been used in the two sets of elastic control when R > 1, as noted earlier (2.1).
experiments. When the strain rates were the same, the results of In general, buckling in viscoelastic media involves consideration
using the two techniques were basically identical (Zhang et al., of relaxation times of both layer and host, and the situation may be
2000; Jeng et al., 2002). complicated (Mühlhaus et al., 1998; Jeng and Huang, 2008). Jeng
Numerical results are consistent with theoretically predicted et al. (2002, 2008) demonstrated that a waveform of attenuated
initial growth rates, which depend on wavelength/thickness of amplitude may develop in viscoelastic media at a stress level below
harmonic components, either individually or combined in an the threshold stress for buckling in an elastic layer and with pre-
imposed initial amplitude spectrum. The results are most clear for dicted and observed wavelengths greater than the dominant
linearly viscous layer and matrix. If a low-amplitude single wave- wavelength. Numerical models using viscoelastic layer in a visco-
form is initially present, this will amplify at the theoretically pre- elastic matrix have been run by Zhang et al. (2000), restricting study
dicted rate, whether or not the waveform is that of the dominant to cases in which the viscous and elastic dominant wavelengths are
wavelength (Mancktelow, 1999). With initial random perturba- the same. As for a viscoelastic layer in viscous matrix case, there is
tions, quasi-periodic folds develop with average wavelength close a transition from viscous properties controlling buckling to elastic
to the predicted dominant wavelength (Fig. 10, Fletcher and properties controlling buckling with increasing strain rate, for
Sherwin, 1978; Mancktelow, 1999; Schmalholz and Podladchikov, a given set of viscous and elastic parameters (Zhang et al., 2000). In
1999, Fig. 3), provided that the amplitudes of the initial irregular- one example studied by Mancktelow (1999) and Zhang et al. (2000)
ities are sufficiently small. The geometry of the final folds does, the initial layer configuration was in the form of a sinusoidal
however, reflect the particular distribution of initial irregularities perturbation with wavelength less than that of the theoretical
(e.g. Fig. 10, Mancktelow, 1999, Fig. 17). In non-linear viscous dominant wavelength for the cases mL/mM ¼ EL/EM ¼ 20, 50, 100, and
materials, growth rates are enhanced but the final fold form bears 200. At low strain rates the input waveform amplified at a rate
a similar relationship to the initial irregularities as for Newtonian consistent with viscous theory, but without the development of the
materials (e.g. Mancktelow, 1999, Fig. 18). dominant wavelength. At high strain rates, elastic effects became
In nature, folds sometimes occur in packets, and various authors dominant and folds at close to the dominant wavelength developed,
have investigated the growth of folds from an initial isolated replacing the imposed initial waveform, although this change in
perturbation, typically of bell shape (Biot, 1961; Cobbold, 1975, wavelength does not occur if the initial waveform is perfectly
1977; Abbassi and Mancktelow, 1990) to study how such packets sinusoidal and symmetry is maintained at the model boundaries e
might develop. The phenomenon is referred to as serial folding, and in other words in such a case the folds that develop are metastable
it is observed that folds develop sequentially outwards from the (Zhang et al., 2000). Slow strain rates led to localized growth of folds
location of the initial perturbation. The wavelength of these folds is from a single isolated initial perturbation, as expected for a viscous
close to that predicted by theory for the dominant wavelength (Biot layer and matrix, whereas fast strain rates led to the development of
et al., 1961). With enough strain, a packet of high amplitude periodic folds along the layer, with little difference in amplitude or
localized folds may develop (Mancktelow, 1999, Fig. 12). Some wavelength between the fold sited on the initial perturbation and
authors consider non-linear effects to be important in under- the rest of the folds in the train (Fig. 11).
standing serial folding (Hunt et al., 1996, 1997, 2006). The extent to which elastic properties are significant in folding
If the layers have both viscous and elastic properties, the situation under natural conditions is unclear. For small folds in competent
is more complicated than when considering viscous properties layers in quartz or calcite veins in slates or schists or quartz-feldspar
alone. The behavior of the system then depends not just on veins in granitic rocks, which are common in the internal parts of
viscosity contrast, as for viscous layers, but also on the imposed orogenic belts, deformed at typical rates of 1014 s1, it seems that
layer-parallel stress or strain rate. Perhaps the Maxwell model is viscous properties will control fold growth. The value of the strain
the most realistic and yet simple constitutive model that has both rate employed in the models of Mancktelow (1999) and Zhang et al.
viscous and elastic properties and that is a good analog for rocks. (1996, 2000) that resulted in elastic control of buckling was
1010 s1. Such a high value may occur locally, such as in narrow
shear zones, but it is unlikely to be met in most parts of orogenic
a belts where most natural folds develop. It should be added, however,
that high values of mL/mM also favour an elastic response to buckling

b
a
c

Fig. 10. Examples of single-layer folds produced in numerical models with interfaces b
possessing random initial irregularities. (a) Linear viscous materials with mL/mM ¼ 50
and with initial layer length 15 times the dominant wavelength, and maximum
amplitude of irregularities about 0.02 of initial layer thickness, at about 30% shortening
(Schmalholz, 2006, Fig. 5a). (b) Viscoelastic layer and matrix, but with fold growth Fig. 11. Numerical models of single layer folds in viscoelastic media developed from an
dominated by the viscous properties, with mL/mM ¼ 50, and maximum amplitude of isolated perturbation to show the effect of strain rate on fold development. Initial
initial irregularities 0.05 of initial layer thickness at 55% shortening (Mancktelow, 1999, configuration and rheological properties the same for the two models. EL ¼ 3.5e11,
Fig. 17). (c) The same as (b) but with maximum amplitude of initial irregularities 0.005 Poisson’s ratio ¼ 0.25, mL ¼ 2e21, Deborah number ¼ me/G _ ¼ 0.00007 for top case and
of layer thickness (Mancktelow, 1999, Fig. 17). 0.7 for lower case for both layer and matrix (Zhang et al., 2000, Fig. 7).
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2049

(Schmalholz and Podladchikov (1999), Fig. 2b), and conditions that they do in an isotropic matrix, but rather propagate well into the
would most likely lead to an elastic response to buckling would matrix (Kocher et al., 2006, 2008). Numerical models show that in
therefore be a combination of fast strain rate and high mL/mM. a Newtonian matrix the folds maintain a uniform fold form of
In numerical models where the competent layer is strongly non- chevron style, whereas in a power-law matrix the fold form
linear or plastic, the tendency is for fold growth to be localized and changes and the style is a mixture of kink and chevron (Fig. 12).
the lateral propagation of folds away from the initial perturbation is
reduced (Zhang et al., 1996, Fig. 6; Mancktelow, 1999, Fig. 11). 2.4. Multilayer folding: theory and modeling
A fuller discussion of analog and numerical modeling of folds with
non-linear and power-law rheology, and the rheological effects Theoretical treatments of multilayer folding follow generally
on fold growth and geometry, will be addressed in subsequent a similar approach to those for single layers, except that in this case
sections. multiple interfaces and layers of several different viscosities are
Anisotropy is an important property of many rocks and, involved (Biot, 1961; Currie et al., 1962; Ramberg, 1964, 1970;
considering the common occurrence of folds in quartz/calcite/ Johnson and Fletcher, 1994). Many different configurations have
feldspar veins in schists, anisotropy in the matrix is probably much been considered, but it is useful to consider several particular types,
more important than anisotropy in the stiff layer in influencing the which are illustrated in Fig. 13 (with some natural examples shown
buckling instability. A characteristic of folding in this situation is in Fig. 2). If several stiff layers of a given viscosity and constant
that the folds do not die out quickly away from the stiff layer, as thickness are separated by layers of low viscosity, the response of

a b

c d

Fig. 12. Numerical models of folding of a single Newtonian layer embedded in a matrix of various properties (Kocher et al., 2006). The same initial random layer irregularities (red
noise) and the same ratio of normal viscosity of layer to normal viscosity of matrix, mL/mM ¼ 50, in all cases. All shortened by 40%. (a) Linear isotropic viscous matrix; (b) isotropic
power law (nM ¼ 3) viscous matrix; (c) anisotropic (dM ¼ 6) linear viscous matrix; (d) anisotropic (dM ¼ 6) power-law (nL ¼ 3) viscous matrix.
2050 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

a b e
Ι
Ι
C
Ι Ι
C
C

C
C Ι
Ι
C Ι
C
C
Ι
Ι
Effective single layer True multilayer Internal - rigid confinement

c d f
Ι 1
C M

Ι 2 M

C
Ι 3 M

C
Ι M

C
Ι M 4
Internal - self confinement
Independent Independent - different viscosities
Fig. 13. Idealized configurations of multilayer buckling. (a) Package behaves as an effectively anisotropic single layer. (b) Buckling instability in individual layers coupled to those of
neighbors (“true multilayer” of Schmid and Podladchikov, 2006). (c) Stiff layers of the same composition folding independently of neighbors. (d) Fold wavelength of individual layers
determined by local viscosity contrast, m4 > m2 > m3 > m1 > mM (after Ramberg, 1964). (e) Internal buckling of laminated medium, equivalent to an anisotropic layer sandwiched
between two stiff layers (after Biot, 1964). (f) Internal buckling of laminated medium equivalent to an anisotropic self-confined medium (after Biot, 1965c).

the system depends on the number and spacing of the stiff layers as (1994) carried out a thick plate analysis of multilayer folding
well as the viscosity contrast between the stiff layers and the for a system consisting of layers of arbitrary thicknesses and
matrix. If a multilayer of thickness H, embedded in an infinite viscosities that allows accurate calculation of growth rates and
matrix of viscosity mM and consisting of alternating stiffer and softer dominant wavelengths by an iterative process. Depending on the
layers of viscosity m1 and m2 occupying fractions a1 and a2 of the layer configuration, this may result in growth rate spectra with
total thickness, it can be considered an effectively anisotropic single more than one maximum, which is the situation investigated to
layer, with normal and shear viscosities, mn and ms given by: find conditions that may result in parasitic folding (Ramberg,
1963, 1970; Johnson and Fletcher, 1994, chapter 6; Frehner and
m1 m2
mn ¼ a1 m1 þ a2 m2 ; ms ¼ ; (10) Schmalholz, 2006; Treagus and Fletcher, 2009) discussed
a1 m2 þ a2 m1 below. For any particular configuration, change of the preferred
If mM/ms << 0.2(ms/mn)1/2, a relationship for the dominant wavelength with shortening can be found as for single layers (see
wavelength identical to Eq. (1) can be derived (Biot, 1965b), if h is Eqs. (7), (8)), and is expressed for multilayers, as for single layers,
replaced by H and mL by mn in equation (1). by an expression equivalent to Eq. (5) (Johnson and Fletcher,
1994, p. 312).


ld mn 1=3 As spacing between stiff (competent) layers is increased, in
¼ 2p : (11) a simple package with stiff layers of equal thickness and spacing,
H 6mM
there is a transition from effective single layer behavior (Fig. 13a)
It corresponds to the situation shown in Fig. 13a. Biot called this to what Schmid and Podladchikov (2006) call “true multilayer”
similar folding of the first kind, although this is not a similar (Class 2) behavior (Fig. 13b) and then a transition to single layer behavior,
fold according to the standard geometric classification (Ramsay, in which the layers fold independently (Fig. 13c, and a natural
1967, p. 367), but rather a parallel fold (Class 1B). example in Fig. 2a). As the spacing is progressively increased,
If mM/ms >> 0.2(ms/mn)1/2, the bending resistance of the indi- growth rates increase over what is predicted by the Biot theory
vidual stiff layers in the multilayer becomes significant, and the for a single layer, reaching a maximum when the alternating stiff
response of the multilayer corresponds to the situation in Fig. 13b. and soft layers have equal thickness (Fig. 14), which corresponds
The dominant wavelength is given by: to true multilayer behavior (Fig. 13b). Growth rates also increase
as the number of layers is increased, but tend towards a ‘satura-

 tion’ value that depends on viscosity ratio (Schmid and
ld N mL 1=3
¼ 2p ; (12) Podladchikov, 2006, Fig. 5). As argued by Ramberg (1960, 1961),
h1 6mM
the key to determining whether the package behaves as a true
where h1 is the thickness of the stiff layers within the multilayer multilayer or as independent single layers is the spacing in
and N the number of stiff layers. Biot called this similar folding of the relation to the dominant wavelength. When spacing is greater
second kind, and in this case the overall geometry of the folds is than the single layer dominant wavelength, ld, a distance that
similar (Class 2), achieved by alternating Class 1B and Class 3 Ramberg called the zone of contact strain, the individual layers in
geometries of the stiff and soft layers. the package behave independently. Schmid and Podladchikov
The above expressions for dominant wavelength are derived (2006) argue in a reciprocal sense that when the spacing is less
from thin plate analysis (Biot, 1961, 1965b). If shear stresses than 1/ld the package behaves as an effective single layer. These
between layers are significant or if there are intervening soft distances correlate fairly well with the ends of the central plateau
layers, thin-plate theory is inadequate. Johnson and Fletcher in Fig. 14a.
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2051

a b

Fig. 14. Theoretical dominant wavelengths (a) and corresponding growth rates (b) for folds in individual layers in a multilayer package consisting of N stiff layers of viscosity mL,
thickness h and spacing s separated by layers of viscosity, mM, with the whole package embedded in two half spaces of viscosity, mM$mL/mM ¼ 250. Growth rates and wavelengths are
normalized against the Biot thin-plate values (from Schmid and Podladchikov, 2006, Fig. 3).

Treagus and Fletcher (2009) examined scales of folding in symmetric to asymmetric minor folds around a major fold with
multilayers, specifically asking why small-scale folds initiate in progressive deformation. Frehner and Schmalholz (2006) provide
multilayers, when it is known from theory and modeling (Biot, another model whereby minor folds develop and grow more
1961, 1965a; Ramberg, 1963, 1964; Johnson and Fletcher, 1994; quickly than major folds; here, the important mechanism is that the
Mühlhaus et al., 2002; Schmid and Podladchikov, 2006), that initial layer irregularities play a more significant part in fold initi-
a multilayer comprising numerous stiff layers will fold with ation and growth, the thinner the layer.
a stronger amplification than a single stiff layer in the same host The configuration shown in Fig. 13e in which folding develops in
and that buckling instability increases with the number of stiff a multilayer sandwiched between two very stiff and unfolded
layers. Using analytical mechanical models, Treagus and Fletcher layers was termed internal buckling by Biot (1964, 1965a,b,c). A
(2009) investigate the conditions where small folds of one or two comparable pattern of folding may develop in an infinite multi-
layers might initiate with a strong enough instability to become layer, which Biot (1965c) has termed self confinement (Fig. 13f, with
small-scale ‘minor’ folds preserved around the larger-scale folds. a natural example shown in Fig. 2c). In these cases the wavelength
They find that small folds in one layer are likely to outgrow larger is given by the empirical relation (Biot, 1965c):
multilayered folds (a) if the thin layer is the stiffest of all the layers,

1=6 pffiffiffiffiffiffiffiffiffi
or (b) if the multilayer is narrowly or stiffly confined. Fig. 15 illus- m a
trates the process, and the progressive change from initially
ld ¼ 1:90 1 þ 3:63 1 2 h1 H; (13)
m2 m
where the multilayer consists of N alternating layers of viscosity m1
and m2 and thicknesses h1 and h2, with h1 þ h2 ¼ 2h and a2 ¼ h2/2h
and m1 > m2. The total thickness of the package or wavelength in the
y direction in the case of self confinement is, H ¼ Nh. If the soft layer
is vanishingly thin, a2 ¼ 0 and
pffiffiffiffiffiffiffiffiffi
ld ¼ 1:90 h1 H : (14)
This holds only if there is sufficient lubrication retained at the
interfaces of the layers. This expression also holds approximately
for a multilayer with stiff and soft layers of equal thickness when
m1/m2 < 1000 (see Biot, 1964, Fig. 4).
An analysis of folding has also been carried out for homoge-
neous anisotropic materials, in which discrete layers do not exist.
This corresponds to the conditions required for folding in foliated
rock. Cobbold (1976a, 1976b) showed that symmetric sinusoidal
similar folds of low amplitude in linear viscous anisotropic media
will grow into high amplitude chevron folds if the anisotropy is
large, and Casey and Huggenberger (1985) extended this approach
to treat asymmetric folds in a non-coaxial strain history. Fletcher
(2005) analyzed the instability and development of cylindrical
structures in a power-law anisotropic medium, equivalent to
a foliated rock. He demonstrated the existence of band-like insta-
bilities whose orientation depends on the ratio of foliation-parallel
extension to foliation-parallel shear, but there is no characteristic l/
Fig. 15. Model results of the development of asymmetry in minor folds around
h as there are no layers to give a definition to thickness. Fletcher’s
a larger-scale multilayer fold, where the central thin layer is twice the stiffness of the
outer, thicker, stiff layers. The viscosity ratios of dark layers to matrix are 20 and 40. results (Fletcher, 2005, Fig. 5) include structures that are equivalent
From Treagus and Fletcher (2009). to the self confinement of Biot (1965a).
2052 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

Internal buckling in multilayered and/or anisotropic media, and Kaus and Schmalholz (2006) and Schmid et al. (2008) investi-
the spectrum of folding styles from chevron folds to kink bands, gate numerically the development of folds to finite amplitude in
have been demonstrated in analog models by Ghosh (1968), a three-dimensional strain field, with different rates of shortening
Cobbold et al. (1971) and Latham (1979). Latham (1985a,b) inves- in x and y directions in a competent layer with random initial
tigated relationships across the spectrum of folding, kinking and irregularities in x and y. These studies confirm the analytical finding
faulting, in anisotropic non-linear materials in theory and labora- of Fletcher (1991, 1995) that a waveform with axis (y direction)
tory experiments. Examples of different configurations of multi- perpendicular to the maximum shortening rate (x direction) grows
layer and resulting differences in fold geometry are shown in Fig. 2. most rapidly, regardless of the value of the shortening rate in the
perpendicular direction and that there is no preferred wavelength
2.5. Oblique-layer folding, non-plane strain and non-coaxial in this perpendicular direction. Even so, folds tend to develop with
deformation a characteristic aspect ratio, lx/ly, which depends systematically on
the ratio of strain rates in the x and y directions, with aspect ratio, or
Nearly all the theory and modeling of folds, reviewed above, cylindricity, increasing as strain rate in the y direction changes from
concerns plane strain with layers parallel to the maximum short- negative to positive. Schmid et al. (2008) note that the initial
ening direction and maximum extension perpendicular to the exponential growth rates predicted by theory give way to slower
layering. The waveforms considered are cylindrical, developing into growth rates, controlled by layer length, as occurs in two dimen-
folds with axes parallel to the direction of no strain. It has been sions (Schmalholz and Podladchikov, 2000) and as modified for
shown both theoretically (Fletcher, 1991, 1995) and experimentally, three dimensions by Kaus and Schmalholz (2006).
in single layers (Grujic and Mancktelow, 1997) and multilayers Experiments show that the folds that form under simple shear
(Watkinson, 1975), that folds also develop in plane strain with are, at least for small amplitudes, symmetric (Ghosh, 1966; Manz
maximum extension of the base flow parallel to the layer (rather and Wickham, 1978; Viola and Mancktelow, 2005), as for the
than perpendicular to the layer) and with the fold axes parallel to inclined layers above. Growth to finite amplitudes, both in terms
the direction of maximum extension, In this case the growth rates of growth rate and orientation of the fold axes with respect to the
are reduced, increasing the viscosity ratio required to produce far-field principal strain axes, is complicated by the non-coaxial
significant folding and, for a given viscosity ratio, increasing the nature of the deformation (Flinn, 1962; Ghosh, 1966; Treagus and
amount of layer-parallel shortening that precedes large-amplitude Treagus, 1981). The situation is further complicated for non-
fold growth. Newtonian materials, as the case analyzed exactly by Fletcher
It has been shown (Treagus, 1973) that the same dominant (1995) indicates, since the individual stress and strain rate
wavelength of a cylindrical waveform as given by Eq. (1) would components are related through the second invariant of devia-
arise for Newtonian materials under conditions of plane strain but toric stress, which depends on all three principal stresses. If the
with the layer oblique to the principal compression (but parallel to principal stress in the third dimension (y say, parallel to the fold
the direction of intermediate stress and strain). To first order, axis) dominates the second strain rate invariant, then the
the initial low-amplitude buckles are symmetric (Treagus, 1973; invariant will not change much in response to variations of stress
R.K. Davies and R.C. Fletcher, personal communication, 2009), but component in the xy plane that control fold development. The
become asymmetric during progressive folding, as shown in finite- degree of non-linearity of flow associated with folding and hence
element models (Anthony and Wickham, 1978). The same princi- growth rates of folds will be diminished. In the situation
ples have been shown to apply to folds that develop in non-coaxial described by Eq. (15), if the power-law exponent is 3 and the
deformation, such as in transpression (James and Watkinson, 1994), extension in y is equal to the shortening in x, the effective stress
provided there is a component of shortening along the layer. Folds exponent, n*, becomes 1.2, which is only weakly non-linear. This
in these situations grow perpendicular to the instantaneous effect will not be large in cases for which the fold axis is parallel
maximum shortening direction in the layer. to the intermediate bulk strain and thus to a direction of small
If the layer has initial sinusoidal waveforms in both x and y deviatoric stress.
directions in the plane of the layer, with wavelengths lx and ly,
and with shortening in both x and y directions, Fletcher (1991) 2.6. Folding of oblique layers in general three dimensional strain
shows that a cylindrical waveform (i.e. ly ¼ N) with axis fields
perpendicular to the maximum shortening rate of the basic flow
grows most rapidly, although the growth rate is not markedly Some cases of folds developing in a three-dimensional strain
greater than that when lx/ly > 0, that is when the folds are non- field have been considered in the previous section. If the layering is
cylindrical with increasing aspect ratio as lx/ly increases. This is oblique to all three far-field principal stresses or strain rates, the
true for most values of e_ x =e_ y , but when e_ x ¼ e_ y (that is equal directions of maximum shortening rate in the competent layer
shortening rate in both x and y directions) the instability disap- varies continuously with time and the base strain rate on which
pears and all fold forms grow at an equal rate. If initial random a buckling instability would be superimposed is therefore also
perturbations in both x and y are amplified, the tendency of the changing. This is true for a bulk coaxial strain history, and true but
resulting fold form to cylindricity increases with amplification, further complicated for a bulk non-coaxial strain history. Flinn
and is stronger for a power-law layer than a Newtonian layer (1962) considered how the finite strain and potential positions of
(Fletcher, 1995). In the power-law case, for a cylindrical pertur- fold hinges vary in layers oblique to the principal strain directions
bation with axis parallel to y, the stress exponent n in the in a bulk coaxial strain field of various strain symmetries. Addi-
expression for the growth rate must be replaced by an apparent tional features of folding in 3D will be discussed in 6.4.
stress exponent, n*, where:
 
2 3. Information from wavelength/thickness ratios of single-
4n 1 þ x þ x
* layer folds
n ¼   : (15)
2 2
4 1 þ x þ x þ 3ðn  1Þx
Fold theory and modeling, reviewed in Section 2.1e2.6, provide
where x is the ratio of the in-plane deformation rates of the basic- the basis for analysis and interpretation of natural folds and their
state flow, e_ y =e_ x . wavelengths, leading to information about the rheological
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2053

properties of rocks. Equations, such as (1), (4), (5) and (6), define the growth rate decreases to zero, or in fact becomes slightly
dominant or preferred wavelength-thickness ratios for single-layer negative (Fig. 16). During this stage the growth is essentially kine-
folds under different rheological and strain conditions. For low- matic, although this is only approximately so because the arclength
amplitude folding, these equations can be applied to folding of of the fold is essentially constant, whereas in true kinematic
layers that are not parallel to maximum compression, and to behavior the local change in arclength will depend on position
deformations other than pure shear, such as simple shear and around the fold e shortening in the hinge and extension in the
transpression, with a caveat for non-linear behavior discussed in limbs if the limb dips are greater than 45 .
Section 2.5. Natural and experimentally produced folds show a range in
values of wavelength/thickness or arclength/thickness. If fold trains
are divided up into segments bounded by hinges and/or inflexion
3.1. Wavelength selection
points, the distance between adjacent hinges is a measure of half
wavelength, or the distance between inflexion point and hinge
The irregularities in natural layer interfaces may be considered
a measure of a quarter wavelength (Fig. 17a). Using such a measure,
as the sum of the components of a harmonic series. In linear theory
Schmalholz (2006) shows that the growth of each individual fold in
the growth rates of folds of different wavelength are linearly
a quasi-periodic fold train developed from a layer with random
independent, and there will be competition among the various
initial perturbations follows the normalized amplification curve.
harmonic components. It is assumed that the folds that emerge are
Based on the observation that little change in arclength occurs for
those of the wavelength with the fastest growth rate, that is the
folds with limb dips >15e20 , it has been suggested that the
dominant wavelength or, in the case in which layer-parallel short-
arclength of mature folds provides a good measure of wavelength
ening is significant, those with the wavelength that has the greatest
at the time that wavelength selection ceases (Fig. 4, Hudleston,
cumulative amplification, that is the preferred wavelength. Because,
1973b, 1986; Fletcher and Sherwin, 1978). Frequency distributions
for viscous materials, all harmonic components are amplified, the
of wavelength/thickness for natural folds (Fig. 17b) are similar in
final waveform will reflect the superposition of the amplification
form to amplification spectra, as pointed out by Sherwin and
spectrum on the amplitude spectrum of the initial irregularities
Chapple (1968), leading to the suggestion that the mean value of L/h
(Fletcher and Sherwin, 1978) and thus not be truly periodic unless
for a natural fold population, L=h, may be a good measure of the
the initial amplitude spectrum is periodic.
wavelength/thickness of the folds that have received maximum
The process of wavelength selection, in which a particular
amplification, lp/h. This assumption has been used in estimating
wavelength spectrum emerges and becomes locked in, with hinges
viscosity contrast in natural folds (Sherwin and Chapple, 1968;
fixed, must have ceased by the time the crossover amplitude of
Hudleston, 1973b; Hudleston and Holst, 1984). A test of this
Schmalholz and Podladchikov (2000), discussed in Section 2.2, is
assumption was made by Fletcher and Sherwin (1978) who showed
attained. Schmalholz (2006) finds the limb dip at which the
empirically that this relationship holds provided the initial irreg-
crossover amplitude is reached is about 17, which is in general
ularities in the layer correspond to a spectrum of white roughness,
agreement with earlier studies that noted that most shortening of
which is arguably a reasonable approximation to many natural
the layer ceases when folds attain limb dips of 15e20 (Sherwin
surfaces. If, on the other hand, the harmonic components in the
and Chapple, 1968; Hudleston, 1973a). Schmalholz (2006) shows
initial layer have constant amplitude, L=h underestimates lp/h, by
that if amplitudes are scaled to the crossover amplitude, all plots for
an amount that can be determined and that depends on the
the change in amplitude with strain fall on essentially the same
dispersion of the distribution. In addition, Fletcher and Sherwin
curve (Fig. 16). An additional stage of folding, beyond the point
showed that the dispersion (standard deviation/mean) of the
where the crossover amplitude is exceeded, can be identified when
frequency distribution of L/h is related to the relative bandwidth of
the amplitude spectrum of the harmonics present in the fold shape.
The relative bandwidth can then be used to estimate the maximum
amplification the folds have undergone.

3.2. Field data and inferences

Data on arclength/thickness, L/h, measured in the way illus-

trated in Fig. 17a, are shown for a single population of folds in
Fig. 17b and for published populations of folds in Table 1. Data of
this type have been the subject of recent discussion (Treagus and
Hudleston, 2009; and Schmid et al., 2010). Values of L/h in Table 1
range from 2e30, with means from 3.7 to 14.5 for a variety of
compositions of both layer and matrix. If the mean value of L/h is
taken as the dominant wavelength/thickness, ld/h, and the
viscosity is assumed to be Newtonian, these values imply a range of
viscosity ratios from 2 to 75 applying thin-plate theory and
somewhat smaller values applying thick-plate theory. As Sherwin
and Chapple (1968) and subsequent experiments and numerical
models have shown, these values are so low in most of these
examples that significant layer-parallel shortening would have
occurred before the amplification could have been sufficient to
Fig. 16. Scaled amplification curve (Schmalholz, 2006, Fig. 3), plotting limb dip or bring the folds to the stage at which the wavelengths became
amplitude/wavelength against shortening. The black line is calculated for qo ¼ 15 locked in, that is when the folds attained limb dips of about 15 .
(corresponding to a viscosity contrast of about 50) and Ao/lo ¼ 3e3 (corresponding to
a limb dip at the inflexion point of about 1 ). The fold shapes are numerically simulated
Indeed, for the lowest values of L/h and at the lowest apparent
single-layer folds for a viscosity contrast of 50. The crossover stretch, Sc, is given by viscosity ratios the buckling instability would have been masked by
Schmalholz (2006, Eq. (8)). kinematic amplification.
2054 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

a b

Fig. 17. (a) Local measures of fold amplitude, A, thickness, h, wavelength, l, and arclength, L, for quasi-periodic folds as typically found in nature. These were the measurements used
to produce the histogram in (b) for folds in calcite veins and limestone beds in slates east of Golden, British Columbia (after Hudleston, 1986). M indicates the mean.

If the mean value of L/h is taken as the preferred wavelength/ (1984), and Fletcher’s thick plate theory, if we take the mean
thickness, lp/h, as is appropriate for an initial distribution of L/h ¼ 6.5 to be Lp/h and the amplification required to lock in this
irregularities in the form of white roughness (Fletcher and Sherwin, wavelength as 20, assuming a power-law exponent n ¼ 1 for layer
1978), to attain an appropriate degree of amplification, in the range and matrix (i.e. Newtonian) implies a viscosity ratio of 17 and an
10e100 (Sherwin and Chapple, 1968; Fletcher, 1974; Hudleston and initial layer-parallel shortening of about 35%. The viscosity ratio is
Holst, 1984; Schmid et al., 2010), implies large amounts of short- reasonable, but the measured shortening is probably less than 5%
ening if Newtonian viscosity is assumed. All the studies (except and certainly less than 20%. Increasing the amplification worsens
Holst, 1987) for which strain data are available in Table 1 indicate the problem and reducing it to 5 still requires excessive shortening
limited amounts of shortening, the maximum being T ¼ 1.5 (or (Hudleston and Holst, 1984). If we take a power-law exponent n ¼ 3
shortening of the layer of about 20%), but with likely values being for layer and matrix, with the same amplification, it implies
much less than this. The required amounts of shortening are a viscosity ratio of about 17 and a shortening of 18%, which is within
significantly reduced if power-law behavior is assumed, which the range of estimates of measured strain. Applying different
leads to higher amplification rates at smaller values of ld/h and lp/h analytical solutions to natural single-layer folds, Schmalholz and
(as indicated by Fig. 6). Mancktelow (2008) estimated the viscosity ratio to lie in the
Because of the lack of good knowledge of the initial form and range 20e70 and the power-law exponent of the layer to lie
amplitude of initial layer irregularities, the limited information between 1.8 and 5.
available on the early layer-parallel shortening of folded layers, and A tentative conclusion from the above, and also from the
the possibility of rheological behavior not properly accounted for in discussion of Schmid et al. (2010), is that the measured single-
the theoretical models used in the analysis, it is not possible to layers folds in the rocks shown in Table 1, more closely approximate
make precise estimates of viscosity contrasts and power-law to power-law rheology than to Newtonian or viscoelastic rheology.
exponents from the kind of data presented in Table 1. However, However, it should be pointed out that these data sets of fold
useful information can be obtained from analysis of natural fold wavelengths mostly concern folded veins of quartz, calcite or
populations. As an example, using the data of Hudleston and Holst pegmatite, on the scale of mm to cm. There is very little evidence on

Table 1
Arclength/thickness data for single-layer folds.

Mean L/h Range Number of folds Strain in profile plane, T Stiff Layer Matrix Source
5.5 2e14 473 quartz veins phyllite Sherwin and Chapple (1968)
4.5 2e13 142 quartz veins sandy Sherwin and Chapple (1968)
5.2 582 quartz veins slate Sherwin and Chapple (1968)
6.8 9 quartzite slate Sherwin and Chapple (1968)
5.1 83 quartz veins slate Sherwin and Chapple (1968)
4.0 12 siltstone slate Sherwin and Chapple (1968)
5.7 17 quartzite phyllite Sherwin and Chapple (1968)
3.7 1.5e8 157 quartzo-feldspathic veins pelitic schist Hudleston (1973b)
6.5 29 1.0e1.5 calcite veins slate Hudleston and Holst (1984)
7.0 2e16 233 calcite veins slate Hudleston (1986)
9.4 e 3 1.07e1.27 limestone shale Fletcher (1974)
a e
5.5 4e13 22 quartz veins mafic schist Shimamoto and Hara (1976)
b e
8.0 4e22 52 c. 1.15e1.35 quartz veins psammitic schist Shimamoto and Hara (1976)
c e
8.5 5e29 59 c. 1.15e1.35 quartz veins psammitic schist Shimamoto and Hara (1976)
d e
14.5 10e25 22 c. 1.06e1.20 quartz veins pelitic schist Shimamoto and Hara (1976)
9.2 35 pegmatite granite Ramsay and Huber (1987)
6.2 1.5e17.5 343 2.8e4 quartz layers slate Holst (1987)
7.8 4.8e11.3 4 quartz veins calcarenite Johnson and Fletcher (1994)
12 calc-silicate layer calcite marble Schmid et al. (2010)
6.9 quartz-feldspar pegmatite calcite marble Schmid et al. (2010)
5.9 quartz-feldspar pegmatite felsic gneiss Schmid et al. (2010)
a
These values are modes, not means, and for a subset of the data that includes only folds with limb dips of 75e90 . The mode of this subset will be closer to the mean of the
total fold population than the mean of the subset.
b
Same as above, for a subset with limb dips of 40e55 .
c
Same as above, for a subset with limb dips of 15e30 .
d
Same as above, for a subset with limb dips of 50e60 .
e
These exclude folds outside indicated ranges of limb dips.
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2055

a larger scale, or for a wider range of metasedimentary rock of quantified by employing harmonic analysis (Hudleston, 1973c),
siliciclastic type, in part because true single-layer folds are power functions (Bastida et al., 1999) or other means (Bastida et al.,
uncommon in stratigraphic sequences, especially in the first stage 2005). The departure from sinusoidal shape is predicted by third
of deformation. Multilayer folding is most commonly seen, and order buckling theory (Johnson and Fletcher, 1994) and is clearly
here the wavelength-thickness relationships are harder to analyze. seen in analog and numerical models (e.g., the succession of fold
We can find no convincing evidence, to date, to indicate that a wide shapes in Fig. 16; Hudleston, 1973a).
range of common rocks, such as interbedded quartzites, psammites, Schmalholz and Podladchikov (2000) found that the first two
and pelites, have folded as power-law multilayers. Indeed, it has terms of the Taylor series that can be employed to define the
been argued from other structures such as cleavage refraction and arclength of a sine function, strictly applicable only for small values
conglomerate deformation that suggest quite small viscosity of A/l, match the arclengths of numerically produced single layer
contrasts in rocks (Treagus, 1999; Treagus and Treagus, 2002; Czeck folds up to limb dips of about 50 . The shape of the folds, as distinct
et al., 2009), that under these conditions of deformation, the rocks from arclength, also starts to deviate significantly from sinusoidal
might be considered approximately Newtonian. at about 50 limb dip. Also, as amplitude increases, fold shape
It should be possible to identify buckle folds that develop in increasingly becomes sensitive to non-linearity in the flow law. As
single layers within an anisotropic matrix by the propagation of the power-law exponent of the stiff layer increases, folds tend to be
folds away from the layer well into the matrix, as shown in the sharper hinged and longer limbed, a change associated with
numerical models of Kocher et al. (2006) (Fig. 12). Getting infor- deformation tending to become localized in the hinge region. This
mation on viscosity contrast will be more difficult than for single can be noted by eye if folds of similar arclength/thickness ratio are
layers in an isotropic matrix because yet another parameter, the compared (Fig. 18) and quantified if some appropriate measure of
anisotropy factor, d, is introduced. shape is used. Numerical models by Hudleston and Lan (1994)
Despite the uncertainties discussed above, it should be possible showed, by varying lo/ho, mL/mM, nL and the shape of the initial
for some populations of folds, and using information discussed layer perturbation, that the main factor controlling sharpness of
below in sections 4 and 5, in addition to wavelength/thickness data, fold hinges is nL. They employed a curvature index, Ki (see Fig. 18,
to define with more precision the parameter space in which natural inset), that varies between 0 for a fold represented by a circular arc,
folds lie. For multilayer folds, the eigenvalue/eigenvector method of 1 for a chevron fold, and 0.75 for a sine function with limb dip of
Johnson and Fletcher (1994) offers the possibility of modeling 45 . The biggest shape difference among folds in layers with
specific layer configurations to match natural examples, although
again the number of parameters to consider makes this a challenge.

4. Information from fold shapes

Fold shapes in rocks are highly variable, especially in multilayers

(e.g. the illustrations in Weiss, 1972; Price and Cosgrove, 1990,
chapters. 12, 13; Hudleston and Lan, 1993; Fig. 2). This reflects
differences in the mechanical response due to differences in layer
thickness, viscosity, degree of anisotropy, deformation intensity a
and deformation path, as well as the influence of initial layer
irregularities. Many methods have been proposed for character-
izing fold shape, on the basis of thickness variations (e.g. Ramsay,
1967, p. 360; Lisle, 1997), dip isogons (Ramsay, 1967, p. 363;
Hudleston, 1973c), relationships among selected fold parameters e
including amplitude, wavelength, interlimb angle, hinge curvature
(e.g. Twiss, 1988) e and by utilizing various mathematical func- b
tions, including harmonic functions (e.g. Stabler, 1968; Hudleston,
1973c), power functions (Bastida et al., 1999), conic sections
(Bastida et al., 2005), and Bezier curves (Srivastava and Lisle, 2004).
We do not attempt to review all these here, but refer readers to the
review by Bastida et al. (2005).

4.1. Single layers c

The most regular and reproducible shapes are found in folds that
approach periodic form with a single wavelength, both in nature
(Fig. 1a,b) and in analog or numerical models (Fig. 10). By contrast,
when different wavelengths of comparable amplitude are repre-
sented in finite amplitude folds e which is most likely to occur
when the initial irregularities are large and the selectivity of the
folding poorethe shapes are less regular (e.g. the right hand side of
Fig. 18. Numerical models of single-layer folds in a power-law stiff layer embedded in
Fig. 10). Periodic buckle folds are initiated with close to sinusoidal a Newtonian matrix, with base viscosity contrast mL/mM ¼ 100, developed from an
form and these grow to become more rounded in the hinge than identical starting sinusoidal shape with Lo/ho ¼ 12 and Ao/ho ¼ 0.1. Three cases shown
sinusoidal, unless the layer rheology is strongly non-linear for nL ¼ 1, 3 and 10, all at 50% bulk shortening. Also shown are the sharpness
(Chapple, 1968), strongly anisotropic (Lan and Hudleston, 1996), parameter, Ki, and schematically the strain in the inner and outer arcs of the fold,
separated by a dashed line representing the finite neutral surface. The inset shows how
strain softens (Neurath and Smith, 1982), or fracturing occurs curvature index is defined for a fold segment between hinge and inflexion point (see
(Tentler, 2001) in the hinge, in which case the fold shape can also Hudleston and Lan, 1994). K is curvature. The value of Ki for a sine function with
become quite angular. The change in shape with fold growth can be the same limb dip (about 63 ) for these folds is 0.86.
2056 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

different power-law exponents is for wavelengths greater than

about 10 and at limb dips of 40e70 (Fig. 18). If the shapes of folds at
a given limb dip are compared, hinge sharpness, as measured by Ki,
increases with L/h with little dependence on nL, until values of
about L/h z 10 are attained (Fig. 19). For values of L/h greater than
about 10, hinge sharpness changes little with increase in L/h, but
there is a clear increase in sharpness as nL is increased. An alterative
plot of hinge sharpness at a given value of L/h (with L/h > 10) gives
similar results and shows that hinge sharpness increases only
slowly with limb dip (Fig. 20). This dependence of hinge sharpness
on non-linearity of flow behavior in the stiff layer is expected and
consistent with Chapple’s (1968) theoretical results of folding
a viscous-plastic beam in which plastic failure occurs in the hinge.
In principle, hinge sharpness can be used to gain some infor-
mation about rheological properties from natural folds (Hudleston
and Lan, 1993). Data for a set of minor folds in siltstone layers in
shale from the Appalachians are included in the plots in Figs. 19
and 20. The data cluster above the line representing nL ¼ 3, with Fig. 20. Sharpness parameter, Ki, as a function of limb dip at fixed arclength/thickness
ratio, L/h z 12, for numerical (solid lines) and natural and analog (symbols) folds. Data
many points above the line for nL ¼ 10. This suggests highly non- from Gairola and Kern (1984) are for experimentally-produced folds in limestone.
linear rheological behavior, although, as discussed below, other Otherwise as in Fig. 19. From Lan and Hudleston (1995a).
factors can also lead to sharp-hinged folds. It is clear from

numerical simulations that random irregularities in the initial

layer lead to fold trains with considerable variation in the shapes
a of individual folds (Fig. 10, Schmalholz and Podladchikov, 2001,
Fig. 8, 11; Schmalholz, 2006, Fig. 5; Mancktelow, 2001, Fig. 7).
Schmalholz (2006), however, shows that individual folds of
different amplitude and somewhat different shape in a fold train
follow almost identical scaled amplification curves. This demon-
stration and the results of Hudleston and Lan (1994) support the
proposition that the folds in a fold train that have single well-
defined hinges but vary in amplitude and L/h should have values of
relative hinge curvature that reflect the stage of fold growth and
the degree of non-linearity in the flow law. A limitation of using
relative curvature to identify non-linear rheological behavior is
that a value of L/h ¼ 10 lies towards the upper end of the range of
mean values reported for studied natural fold trains (see Table 1).
The folds used in shape analysis, however, do not have to be ones
corresponding to the dominant (or preferred) wavelength and
many natural fold populations, even those with low mean values
of L/h, contain folds with L/h  10.
Another factor influencing the sharpness of folds is anisotropy.
b Fig. 21 compares fold shapes in single layers in which both layer and
matrix are anisotropic, with different degrees of anisotropy, d. It
should be possible to identify situations in which anisotropy has
influenced folding by the fabric (layer-parallel cleavage or schis-
tosity) in the rocks and by the development of layer-parallel shear
strain in the limbs of the fold; such strain being minimal in isotropic
layers. This can be seen by comparing the deflection of layer normal
markers in Fig. 22a and b. Effective layer anisotropy can also be
achieved by alternating stiff and soft isotropic layers within an
isolated layer package that may behave as a composite single layer
during buckling (Fig. 22c). Sharp-hinged folds of similar shape
(measured by curvature index, Ki) and formed in three different
ways are compared in Fig. 22. This serves to demonstrate that layer
shape alone is not sufficient to distinguish these situations; it
should be considered in combination with internal strain distri-
bution and fabric pattern.
Fig. 19. Sharpness parameter, Ki, as a function of arclength/thickness, L/h, at fixed limb
It seems unlikely that the effects of elasticity will be manifest in
dip (55 ) for numerical (small crosses joined by solid lines) and natural and analog most natural small folds (see Section 2.6), and there is conflicting
(symbols) folds: (a) at a limb dip of 55 ; (b) at a limb dip of 70 . Selected data for information on how elastic versus viscous properties affects the
analog models are from Hudleston (1973a,b,c) and Abbassi and Mancktelow (1999); regularity of fold trains. In numerical models of Schmalholz and
also data for Cruikshank and Johnson’s (1993) numerical model. Solid line data are
Podladchikov (2001, Fig. 11), there is much greater variation in
interpolated from results of numerical models with Lo/ho values of 6e30, nL ¼ 1, 3, 10
and mL/mM ¼ 100. Appalachian data are for folds in siltstones in shales. From Lan and shape and strain associated with individual folds in layers in which
Hudleston (1995a). elastic properties influence folding than in purely viscous layers. By
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2057

a
a

b
b

c
c
Fig. 21. Numerical models of single-layer buckle folds to show the effect on fold shape
of increasing anisotropy, as measured by d ¼ normal viscosity/shear viscosity. In all
cases Lo/ho ¼ 12, Ao/ho ¼ 0.1, nL ¼ 1, mL/mM ¼ 100 and bulk shortening ¼ 40%. The value Fig. 22. Numerical models of single layer (a,b) or effective single layer (c) buckle folds
of the curvature index, Ki, is shown beside each fold. (a) dL ¼ dM ¼ 1 (isotropic layer and of similar values of L/h and showing similar degrees of angularity (sharp hinges and
matrix), (b,c) dL ¼ dM ¼ 20 and 50 (anisotropic layer and matrix). After Lan and straight limbs) as measured by shape index, Ki, produced by different combinations of
Hudleston (1996). rheological conditions. (a) In a highly non-linear isotropic stiff layer with dL ¼ 1,
nL ¼ 10; (b) in a strongly anisotropic layer with dL ¼ 15, nL ¼ 1; (c) in a composite layer
consisting of three isotropic stiff layers and two soft layers, with the soft layers the
same as the matrix. In all three cases Lo/ho ¼ 12, Ao/ho ¼ 0.1, mL/mM ¼ 100 and
shortening ¼ 40%. In (c) the effective anisotropy of the package, deff ¼ 24, is based on
contrast, in models of Zhang et al. (2000), there is more variation in
Biot’s theory (Biot, 1965a, p. 432e433.). The matrix is Newtonian and isotropic in each
fold shape in layers controlled by viscous properties than in layers case (i.e. nM ¼ 1, dM ¼ 1). (Based on Fig. 6 in Lan and Hudleston, 1996). The value of
in which folds are controlled by the elastic properties. Differences curvature index, Ki, is shown beside each fold.
in matrix rheology and choice of elastic and viscous dominant
wavelengths likely explain these results. 4.2. Multilayers
The shape of buckled layers in 3D e in part the degree of
cylindricity of the folds - may contain information about the Fold shape is much more variable in multilayers than in single
rheology of single layer folds. Fletcher (1995) showed that the layers (e.g. Fig. 2). Characteristic shapes include sinusoidal,
selectivity of buckling for harmonic components with large aspect chevron, kinks e with axial surfaces inclined to the average
ratio in x and y directions in the plane of the layer (with longest orientation of the layering e and conjugate (or box) folds. Straight-
dimension perpendicular to the shortening direction) in plane limbed folds, either of chevron or kink style, imply effective
strain depends on the power-law exponent, nL, such that folds of anisotropy, but do not require discrete layers (as in the case of kink
greater cylindricity will develop from a layer with random pertur- bands in slate) or brittle failure in the hinges. Chevron folds (Fig. 2d)
bations in both x and y directions in the layer interface for highly can develop in several ways: from initial sinusoidal shapes (e.g.
non-linear rheology (very large value of nL, equivalent to plastic Johnson, 1977), by the appression of conjugate folds (see Price and
behavior) than for Newtonian rheology (Fig. 23). Viscosity contrast Cosgrove, 1990, Fig. 13.45) or by the intersection of kink bands
also has an effect, with high contrast enhancing cylindricity, but (Paterson and Weiss, 1966; Cobbold et al., 1971). For symmetric
only for non-linear rheology. folds, a chevron style is inherently favored by anisotropy (Bayly,
Information about folding mechanisms by which the strain in 1974; Cobbold, 1976a and discussion above). A discussion of the
the competent layer is accommodated can be found by applying the various styles and significances of folds in anisotropic rocks is given
techniques of Bobillo-Ares et al. (2004). Fold shapes can be by Price and Cosgrove (1990).
matched by trial and error application of their FoldModeler Bayly (1970, 1974) proposed methods for deriving information
program (Fig. 24), if information on cleavage orientation around the on viscosity contrast and anisotropy from the geometry of chevron
fold is available and the assumption is made that cleavage reflects folds in bilaminate materials, by considering relative hinge thick-
the XY plane of the strain ellipsoid (with X  Y  Z being the lengths ening and change in limb length (Bayly, 1970) and by estimates of
of the semi axes of the ellipsoid). Unless the value of strain is known the energy consumed in incremental shortening of chevron folds
at some point in the fold, a full model of fold development cannot with distinct hinge regions and straight limb regions (Bayly, 1974).
be made. This method assumes the processes operating to produce Many assumptions, including Newtonian viscosity, are involved in
the observed shape and strain are combinations of flexural flow, applying these methods and they have not been subsequently
tangential longitudinal strain and homogeneous flattening. applied or tested in any significant way.
2058 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

a b

Fig. 23. Plan view in the plane of the layering showing the effects of selective amplification of initial, random perturbations in the folding of an isolated viscous layer (nL ¼ 1.001) in
a viscous matrix with viscosity ratio, mL/mM ¼ 100 after amplification of 100. Contours are normalized by the maximum height of the surface above its mid plane. The length of the
side of the square region is 10 ld. (a) Newtonian layer, (b) power-law layer with nL ¼ 104 (Figs. 2D and 3D in Fletcher, 1995).

Chevron folds tend to develop in a multilayered sequence with mechanical properties than has so far been achieved, but the fact
stiff layers of similar thickness separated by thinner soft layers. that folds of a given shape can be arrived at by different processes
Ramsay (1974) showed how characteristic modifications to the and that layers of varying thickness and rheological properties may
shape of the folds in the hinges, or accommodation structures, make up many multilayer fold packages makes realizing this
based on geometrical considerations and the assumption that the difficult.
stiff layers tend to maintain constant length and thickness, were
required if individual unusually thick layers are introduced. The 5. Strain measurement from folds
typical accommodation structure is a keel-shaped or bulbous hinge
zone (Ramsay, 1974, Fig. 8). This structure can also be produced 5.1. Elementary buckle shortening
when an unusual amount of slip is developed on the fold limbs in
individual bedding planes (Price and Cosgrove, 1990, p. 319e321). For rock layers that have folded by buckling, the most elemen-
The large variation in shape of multilayer folds suggests that tary calculation of strain assumes that all the layer shortening is
shape may potentially provide much more information on expressed in buckling. The bulk shortening stretch is the fold

c
a
b

e f g

Fig. 24. Example of fit of a natural fold (not shown) by superposition of strain patterns applying the FoldModeler technique of Bobillo-Ares et al. (2004). (a) Ramsay’s normalized
thicknessedip (t0 q vs q) classification of theoretical folded layer (line) and natural fold (dots). (b) Initial configuration of theoretical layer. (c) Folded configuration of theoretical layer
showing strain ellipses with long axis orientation, assumed to be parallel with cleavage. (d) Comparison of values of parameters of the modeled fold with corresponding values for
the natural fold (aspect ratio is final amplitude/final width of the central (generating) fold surface; to is thickness at the hinge, A is the amplitude of the outer arc of the final fold. (e)
and (f) Maximum stretch direction vs dip (4eq) diagrams for the outer and inner arcs respectively showing data obtained for the modeled folded layer (line) and natural fold (dots).
(g) Strain ratio vs dip (S1/S2 vs q) diagram for outer and inner arcs showing strain pattern predicted for theoretical fold (all data from Fig. 13, Bobillo-Ares et al., 2004).
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2059

‘wavelength’, or series of wavelengths (recognizing that quasi- 5.3. Shortening in analog and numerical models
periodic or non-periodic folds do not have wavelengths in the
mathematical sense) (Fig. 4c or 17a) divided by the arclength, the In an earlier review (Treagus, 1997), compilations of bulk model
same fold length measured around the folded layer. This method shortening versus fold limb dip in analog and numerical models
has been employed since some of the earliest studies of fold belts, (Fig. 26), sourced from a wide range of published fold studies,
including those of Claypole (1885), and Chamberlin (1910) in the showed that there is commonly between w5 and 20% bulk (layer-
Appalachians. The same principle is employed in line-length parallel) shortening before buckles start to amplify measurably (at
conservation in producing balanced and palinspastically restored around 10 limb dip), with the greater shortening for models with
cross sections in fold-and-thrust belts (Dahlstrom, 1969). lower viscosity ratios. A comparison of Figs. 25 and 26a reveals that
For one fold, with wavelength, l, and the around-fold arclength, in the latter, the analog model with viscosity ratio ¼ 100
L, the shortening strain, e (-ve) and stretch, S (<1), are defined (Hudleston, 1973a; Hudleston and Stephansson, 1973) is almost
(Fig. 17): identical to curve S (sinusoid) in Fig. 25, if its origin is shifted to the
0.9 point (10% shortening). However, the numerical models in
S ¼ ð1 þ eÞ ¼ l=L: (16) Fig. 26b show curves with a straighter trend at higher limb dip, as
Measurements of l and L must be made in a section perpen- did several multilayer models not shown here.
dicular to the fold hinges, and the assumption is that there is no
displacement gradient in the direction of the hinges. 5.4. Strain contour map of Schmalholz and Podladchikov

Schmalholz and Podladchikov (2001) provide a new method for

estimating strain and viscosity contrast from fold shape, based on
viscous and viscoelastic finite-amplitude fold modeling. The term,
5.2. Shortening according to fold shape fold shape, does not here mean shape analysis as considered in
section 4, nor the types of shape shown in Fig. 25. Instead, it
The method above assumes that all the layer shortening is concerns amplitude/wavelength (A/l) and layer thickness/wave-
expressed in buckling: i.e. there is no internal layer shortening length (h/l) measurements plotted on a strain contour map
preceding or accompanying folding. With this caveat, it is possible (Schmalholz and Podladchikov, 2001, Fig. 6), as shown in Fig. 27.
to determine the amount of shortening for folds of different shapes, The theory and modeling behind this plot have been reviewed in
such as sinusoids, circular arcs and types of angular folds (Currie section 2 above. For any natural or modeled fold where A/l and h/l
et al., 1962, Fig. 7; Johnson, 1970, Fig. 4.3; Treagus, 1997, Fig. 19.2), can be measured, estimates of bulk strain (% shortening) and
as shown in Fig. 25. The relationships between stretch, S, and limb viscosity ratio (layer/matrix) can be read from the contour lines
dip, a, are very simple and explicit for circular and chevron folds, in Fig. 27. The strain is measured from the point at which the
but for a sinusoid of the form y ¼ m sin x, the stretch, S, cannot be nucleation amplitude (see Section 2.2) is reached. It will be an
expressed explicitly and involves an elliptic integral (Treagus, underestimate if the initial amplitude is less than the nucleation
1997). amplitude and an overestimate if it is greater. In addition it is
These purely geometric approaches are useful in providing assumed that the fold initiated at the dominant wavelength. If it did
approximate or minimum bulk shortening for folds of a particular not, the growth rate will be less than that assumed for the plot and
shape (Fig. 25). Pure buckling folds in single layers might be thus the strain derived from it an underestimate. These underesti-
expected to follow the sinusoid curves more closely than the mations or overestimations are likely to be small. Schmalholz and
circular or chevron shapes (as demonstrated by Schmalholz and Podladchikov test their method against numerical and analog
Podladchikov, 1999, for dips up to 50 ) and these curves might be experiments in which the actual strain is known, and illustrate it with
a good approximation for thin stiff layers with viscosity contrasts of two examples of folded layers pictured in Ramsay and Huber (1987,
100. Fig. 19.11) and Weiss (1972, plate 171). The A/l and h/l values for

a b

Fig. 25. (a) After Currie et al. (1962, Fig. 7). Amplitude/half-wavelength, 2A/l, and limb dip, q, vs shortening for folds of different geometry - circular, sine, sawtooth and box. (b)
Fig. 19.2 from Treagus (1997), of limb dip vs stretch for circular, sinusoidal and chevron folds.
2060 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

a b

Fig. 26. (a) Compilation of data on changes of limb dip with shortening for analog models, from Treagus (1997, Fig. 19.4), with original data from Hudleston and Stephansson (1973)
(solid lines e single-layer folds), Treagus (1972) (light shaded area e stiffer single layer folds with L/h of 11e15), (dark shaded area e softer single layer folds with L/h of 6e9). M1, M2
multilayers from Treagus (1972) (viscoplastic materials) and Johnson (1977) (elastic materials) respectively. (b) Compilation of data on limb dip vs shortening for single-layer folds
developed in theoretical and numerical models by various authors, as indicated, from Treagus (1997, Fig. 19.5a). Numbers by each curve in (a) and (b) indicate viscosity ratio, mL/mM.

the former, given in Schmalholz and Podladchikov (2001, Table 2), initial layer-parallel shortening, active buckling, and late fold ‘flat-
lead to a strain determination of 70e72% shortening for this fold tening’, what distinguishes Schmalholz’s (2006) approach is that
train; a significant shortening. the three stages are precisely defined (Fig. 16). Only the first stage
has a variable degree of strain, which is related to the viscosity ratio
5.5. Scaled stretch and amplification, and stages of folding and the initial amplitude. As reviewed earlier, Schmalholz presents
a scaled amplification equation, which allows the general amplifica-
Schmalholz (2006) provides another method of determining tion curve to be drawn, as in Fig. 16. This approach derives from
strain associated with stages of single layer folding, but does not earlier papers (Schmalholz and Podladchikov, 2000, 2001) where
measure the whole strain. This paper defines three stages of the crossover strain and crossover amplitude were introduced that
folding, termed nucleation, amplification and kinematic growth. mark the change from the nucleation to amplification stage (see
Although reminiscent of Ghosh’s three stages of buckle folding Sections 2.2, 3.1). Crossover strains may typically range from a few
(Ghosh, 1993, p. 273) or three stages Treagus (1997) described as % (for large viscosity) to about 16% for the example in Fig. 16 (based
on mL/mM z 50) and depend on the amplitude of the sinusoidal
waveforms that make up the initial layer irregularities.
Schmalholz (2006) demonstrates that if the bulk shortening is
normalized or “scaled” to the crossover strain, the 2nd and 3rd fold
stages (amplification, and kinematic growth) have almost identical
forms, for a range of viscosity ratios. However, the results appear to
derive from initial growth rates (given by ao in Schmalholz, 2006,
Fig. 2, equal to qo here) in the range of qo ¼ 10 to 26, that (according
to Schmalholz and Podladchikov (2001), eq. (5), and taking nL ¼ 1)
indicate viscosity ratios of 24e99. There must, therefore, be some
caution in applying Fig. 16 to all single-layer folds, irrespective of
other lines of evidence that might suggest viscosity ratios above or
below this range.
Note that there are some inconsistencies in the definitions and
notations in Schmalholz (2006, pp. 44e45). The scaled stretch (Ss in
Fig. 16) needs to be clarified as the total stretch (S > 1) divided by
the crossover stretch (not crossover strain), which would be better
termed Sc (rather than eC). It might be better to express the stretch
values as <1, i.e. contractional, as these relate more obviously to %
shortening strains associated with folding used by Schmalholz and
Fig. 27. The strain contour map e amplitude/wavelength (A/l) plotted against thick-
ness/wavelength (h/l), from Schmalholz and Podladchikov (2001, Fig. 6), with contours
Podladchikov (2001) and Schmalholz (2006). However, the S > 1
showing bulk shortening and viscosity ratios. See Fig. 28 for an explanation of the values have been retained in Fig. 16, to be faithful to these citations,
black dot. and to follow the precedent set by Sherwin and Chapple (1968) and
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2061

Fletcher (1974), who expressed the strain associated with the initial isolated folds can arise through many mechanisms. Nevertheless,
layer-parallel shortening as T ¼ S1/S2, ¼ 1/S22, in which S2 is the where a layer of rock is disposed in a train of folds of quite regular
stretch in Eq. (16). This convention is also used in Section 2.2. wavelength, such as discussed earlier, it is a reasonable assumption
The amplification stage of folding, defined by the amplitude/ that the primary process is buckling in response to layer short-
wavelength range of 0.05e0.15 (Fig. 16), is represented by a scaled ening. In section 5, we reviewed a number of methods of deter-
stretch from 1 to 1.25, equivalent to 20% shortening for this stage. mining bulk shortening strain from natural buckle folds. In this
Beyond this, for limb dips of >45 , the fold continues to grow section, we will consider what evidence folds might provide to
kinematically. This growth is not truly kinematic, since the analysis indicate the deformation history for the folded layers, on a local or
assumes that the arclength strains uniformly from inflexion to regional scale.
hinge, which will not be the case at high amplitudes, especially at
low viscosity ratios. 6.1. Inferring fold mechanisms from strains and fabric (fanning and
We note that the amplification curve in Fig. 16 departs from the refraction of cleavage)
folding trends for mL/mM ¼ 50 shown in Fig. 26, based on a range of
earlier analog and theoretical viscous models. For folds with limb Two mechanisms for internal deformation associated with
dips of 60 or 70 , Fig. 16 would lead to higher strain estimates than parallel folding have become classics in structural geology, used
those given by the curve in Fig. 26. However, the significance of widely in text books and teaching: flexural flow and tangential
Schmalholz’s new graph (Fig. 16) is that it defines the middle longitudinal strain (Ramsay, 1967, pp. 391e398). These two mech-
amplification stage of folding precisely, from scaled stretch of anisms for accommodating strain in a layer without altering its
1e1.25, regardless of viscosity ratio (within the limits stated above). thickness are shown in Fig. 29. Flexural flow (FF) has strain
Schmalholz (2006) presents an example that combines this concentrated on limbs, and is analogous to bending a stack of paper
‘scaled stretch’ method of strain estimation with the preceding (by sliding). Tangential longitudinal strain (TLS) has strain concen-
Schmalholz and Podladchikov (2001) method, as shown in Fig. 28. trated in the hinges, and is analogous to bending a continuous
The example is a folded quartz vein from Lisle (1995, p. 50). One isotropic material such as rubber. Both fold mechanisms produce
fold has been selected and measured; its A/l ¼ 0.45 and h/l ¼ 0.17 parallel folds (Class 1B, Ramsay, 1967, p. 366), with no change of
values, plotted on the strain contour map (Fig. 27), determine the orthogonal thickness (although there are intra-layer mutually
total shortening strain as 62%. These coordinates place the fold, compensating changes of thickening in the inner arc and thinning
spatially, between the mL/mM ¼ 50 and 250 contours in Fig. 27, as in the outer arc in the case of TLS), and have convergent dip isogons.
indicated by the black dot on that figure. The A/l value for this However, the internal strain trajectories, and the cleavage traces e
example is then used to plot this fold on the scaled stretch graph assuming cleavage is parallel to the XY plane of strain e are
(Fig. 16). This indicates Ss ¼ 2.3, equivalent to 57% shortening significantly different for FF and TLS fold mechanisms. One way of
beyond the nucleation stage (from a scaled stretch value of 1.0 to illustrating the difference is by making a combined isogonestrain
one of 2.3). Hence, for this example (Fig. 28), the three fold stages classification of the folds (Treagus, 1982). Although this plot cannot
emerge as: nucleation, 12% shortening (stretches 0.88 and 1.14) represent the intensity of strain in folds, the orientations of strain
amplification, 20% further shortening (incremental stretches 0.8, trajectories or cleavage traces are sufficient to distinguish certain
1.25) kinematic growth, 46% further shortening (incremental folding mechanism from others, such as FF from TLS. A variation of
stretches 0.54, 1.85). TLS is inner arc collapse, whereby there is no strain in the direction
Multiplication of these three increments of stretch confirms the normal to layering, and all the folding is accommodated by layer-
total contractional stretch of 0.38, i.e. 62% shortening, stated above. parallel shortening strain, associated with removal of material,
most likely by pressure solution (Hudleston and Holst, 1984;
Hudleston and Tabor, 1988). The principal strain orientations are
6. Information on strain and deformation history from
the same as for TLS, but the neutral surface now lies at the outer arc
folded rocks
of the fold.
Some of the earliest applications of finite element (FE) analysis in
Theory and modeling have established that contraction of
structural geology were used in the modeling of folds and revealed
layered rocks will result in buckle folds. However, it does not follow
their deformation patterns. Dieterich (1969) and Dieterich and
that all folds in rocks are a result of simple shortening of layers, and
Carter (1969) first modeled the stress and strain patterns associ-
ated with folding of viscous single layers, and made comparisons
with cleavage patterns in natural folds. Shimamoto and Hara (1976)
used the FE method to examine the geometry and strain in single
layer folds, and their results showed strain trajectories close to
layer-orthogonal in the folded layer, indicating that tangential
longitudinal strain is the dominant mechanism (see Fig. 29). There
have been many subsequent FE studies of folding in single and
multilayer folds (e.g. Lan and Hudleston, 1995b; Hudleston et al.,
1996) that confirm the predominance of the TLS mechanism for
viscous folding models. Hudleston and Holst (1984) analyzed strain
in a folded limestone layer, and concluded that TLS was accompa-
nied by some degree of inner-arc volume loss through pressure
solution. Bobillo-Ares et al. (2000, 2006) considered compatibility
issues associated with the TLS mechanism, and use kinematic and
geometric modeling to examine the developments of strains,
Fig. 28. Example of combining Figs. 16 and 27, after Schmalholz (2006, Fig. 8). The neutral surface and problems of area change. They conclude,
inset drawing and set of measurements are for a fold in a quartz vein taken from Lisle
(1995). This fold is represented by an open circle on this plot and by a black dot on
illustrated by examples, that TLS folding in nature is not perfectly
Fig. 27, where determination of the total strain is made. See text for stretch and strain resolvable without the effects of area/volume changes and other
values obtained for this fold. geometrical effects in hinge regions, echoing some of the effects
2062 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

Finite Neutral Surface

a b
Fig. 29. Illustrations of flexural flow (a) and tangential longitudinal strain folds (b) showing strain ellipses (from Hudleston et al., 1996). The overall shape defined by the inner and
outer arcs is identical in each case.

described for chevron folds (Ramsay, 1974; Bastida et al., 2007), flexural flow as the predominant mechanism of single-layer
considered below. According to Bobillo-Ares et al. (2006), the folding, as well as variations according to scale. Considered
presence of radial wedged quartz veins in competent beds in fold together with the results of Hudleston and Holst (1984) discussed
outer arcs, and bulges or protuberances in inner arcs, are indicative above, folded carbonate layers may exhibit a wide range of strain
of TLS processes in nature. patterns and deformation mechanisms. Our earlier conclusion
Returning to the alternative parallel folding mechanism, flexural (Hudleston et al., 1996) that flexural flow is an unlikely mechanism
flow (Fig. 29), we previously questioned FF as a valid mechanism for for single-layer folding, except in highly anisotropic rocks, would
single layer folding (Hudleston et al., 1996), and concluded from FE appear to be contradicted in some instances in limestones.
models that such a layer would need to have a very high layer- The flexural slip mechanism (Tanner, 1989) is somewhat different
parallel anisotropy (d > 50) to approach the FF mechanism. from the mechanisms of folding discussed above. The essential
However, in a bilaminate multilayer that comprises alternating stiff characteristic is layer-parallel or bedding-plane slip, and flexural
and soft layers, if there are enough layers to represent the material slip folding provides no additional definition of strain within the
as statistically anisotropic, the whole multilayer in bulk might be layers. Tanner’s flexural slip folds are essentially chevron folds. They
considered to have folded according to flexural flow. On a smaller are defined mainly by slip between layers, indicated by geological
scale, the alternating stiff and soft layers in the multilayer would features such as bedding-parallel veins, fractures and duplexes,
fold by different mechanisms: the stiff layers approximately by TLS, plus slickenlines and slickenfibres that provide clear information
as discussed above, with strain and fabrics concentrated in the on slip sense. In contrast, the Ramsay (1974) model for chevron
hinge regions; the softer layers by a combination of flexural flow folding has bonded layers, and deformation by variable amounts of
and strain accommodation; and the boundaries possibly by some flexural flow (FF) within the layers. Our own observations of
degree of (flexural) slip. Fig. 30 shows typical patterns of alter- chevron folds in alternating sedimentary sequences, especially the
nating convergent and divergent cleavage fans, hinge fans and presence of convergent cleavage fans in hinge regions of competent
cleavage refraction that might be expected in folded multilayers of beds and absence of strong strain fabrics on the limbs, not
this kind, according to the strain patterns shown in FE models dissimilar from the patterns in Fig. 30, lead us to conclude that
(Dieterich, 1969; Shimamoto and Hara, 1976), and models of strain there is generally a significant component of tangential longitu-
fanning and refraction (Ramsay, 1967, p. 403; Roberts and dinal strain in competent chevron fold hinge zones.
Strömgård, 1972; Treagus, 1988, 1997). Such patterns (Fig. 30) are Bastida et al. (2007) provide kinematic models of chevron folds
characteristically seen in interbedded sandstones/psammites and using a program, FoldModeler, to find the best combination of
mudstones/pelites, in our experience (e.g. Fig. 31a,b). strain accommodation mechanisms in competent layers that can
The processes of folding in limestone/mudstone alternations, account for the observed strain and fabric patterns in natural
however, may be more complex than shown in Fig. 30. A study of chevron folds. They show that a sequence of different mechanisms
three small folds by Ormand and Hudleston (2003) revealed is required for a good match, and that for angular folds with very
strong curvature and narrow hinge zones, equiareal tangential
longitudinal strain (ETLS) cannot be the exclusive mechanism in
competent beds: other mechanisms and/or area change need to
take place.
Bedding-cleavage In the discussions above, we mainly use strain patterns and
intersection associated fabrics to infer the mechanisms of single or multilayer
folding, but there are other geological structures and features that
provide evidence for the deformation associated with folding.
Veins, slickenlines and fibres and duplex fractures were referred to,
ge above, in terms of evidence for bedding-plane slip and flexural slip
ava
Cle (Tanner, 1989). Another important source of geological evidence is
given by folding of layers that contain oblique markers or surfaces,
such as folded cross-bedding or earlier fabrics (Williams, 1979;
Aller et al., 2010; Bobillo-Ares et al., 2009). These last authors
provide computer modeling of folding of two inclined surfaces by
Fig. 30. Schematic typical cleavage fanning in fold hinges and refraction in limbs
a range of mechanisms including FF and TLS plus ‘flattening’, to
reflecting variations in strain developed in multilayers consisting of alternating stiff seek ‘best fits’ to natural examples. Their two examples of folded
and soft members. sandstone beds, one containing cross bedding, the other an earlier
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2063

Fig. 31. (a) Buckle folds in siltstone layers in slates with cleavage that fans symmetrically about the fold axial surfaces, Boscastle, Cornwall, England; scale in inches. (b) Cleavage
refraction through graded sandstone and mudstone beds on an anticlinal fold limb, Widemouth, Cornwall; hammer head 15 cm. The triangular cleavage fan at the hinge has been
highlighted in white. (c) Flattened buckle folds in quartz-rich layer in schists, near Cap de Creus, Spain; coin 2.5 cm. (d) Largely passive similar folds in amphibolitic gneiss in highly
sheared rocks at the base of a Caledonide basement thrust sheet, Swedish Lapland; compass 8 cm.

fabric, lead to the conclusion that folding was by a combination of A large enough additional flattening or homogeneous pure
mechanisms: flexural flow, tangential longitudinal strain and shear imposed on a parallel (1B) fold would virtually transform it
‘flattening’. into a class 2, or similar fold (Fig. 31d). Similar folds (Ramsay, 1967,
The concept of fold flattening and flattened parallel folds was p. 421) are generally seen in ductile rocks where layering appears to
introduced by Ramsay (1962a, 1967) to explain folds of class 1C have low viscosity contrasts, and deformation is great. They are also
geometry (e.g. Fig. 31c). Theoretical curves for different amounts of seen in salt and ice flows, where similar-style folds are shown by
‘flattening’ (homogeneous pure shear) of an initially parallel (1B) fine marker layers (Hudleston, 1976; Talbot, 1979; Talbot and
fold can be drawn within the 1C field of t0 a and dip-isogon fold Jackson, 1987). Whether such folds are the result of extreme flat-
classification graphs (Ramsay, 1967, Figs. 7e79; Hudleston, 1973c). tening (homogeneous deformation) of weak buckling perturba-
Using examples of ptygmatic folded veins, Hudleston and tions, or are a result of differential simple shear (Ramsay and Lisle,
Stephansson (1973, Fig. 9) revealed that the t0 a plots have close 2000, p. 825) may only be answered by examining their fabrics
fits to flattening curves, but that the values of flattening deforma- and strain patterns. A true axial-planar cleavage would suggest
tion show quite a wide range, for different fold limbs in one fold extreme flattening; divergent cleavages, in particular with asym-
train (strain ratios of 0.2e0.7), indicating inhomogeneous flat- metrical development on either fold limb, suggest differential shear
tening on quite a small scale. In a similar way, Schmalholz and folding.
Podladchikov (2001), noted that not all folds in a layer grow to The fanning and refraction of cleavage, such as the patterns
the nucleation amplitude simultaneously, depending on both initial shown in Fig. 30 that can help to distinguish different mechanisms
amplitude and the initial wavelength. Thus folds in the same fold of folding, can be used in other ways. The sense of cleavage
train can be at different points on the scaled amplification curve at refraction across alternating competent and incompetent layers on
the same time. In addition, for power-law materials the effective folds limbs (e.g. Fig. 31a,b), can be used qualitatively to indicate
viscosity is a function of stress in the layer and this may vary if the relative increases and decreases of effective viscosity. To attempt to
folds grow asynchronously. Local development of serial folds may be quantitative, assumptions need to be made. For example, if it is
relieve stress in neighbouring sections and thus change the effec- assumed that cleavage approximately represents the principal XY
tive viscosity ratio (see Mancktelow, 1999). Folds at low viscosity plane of strain, for both competent and incompetent rocks,
contrasts may thus attain high amplitudes at different times and be cleavage refraction can be taken as a direct indicator of strain
subjected to different amounts of flattening at a late stage. Useful refraction (Treagus, 1983, 1988). However, orientations of cleavage,
methods for estimating the flattening component of strain in alone, do not provide sufficient data to determine strain values, and
initially parallel folds have been proposed by Srivastava and Shah thus are insufficient to determine effective viscosity ratios across
(2006, 2008). boundaries. If, instead, it is assumed that cleavages initiated before
2064 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

folding and are passively deformed bedding-orthogonal marker symmetrical fold waves, and become asymmetric by modification
surfaces (Gray, 1981), the fans and refraction patterns could be used around larger-scale folds, by further shortening and layer-parallel
to determine the strain associated with folding. Treagus (1999) shearing.
provided a possible compromise, and demonstrated that cleavage Treagus and Fletcher (2009) address the question of why small-
traces may be sufficiently close to deformed layer normals, under scale folds initiate in multilayered rocks and are preserved in
the conditions of folding and deformation, to provide a measure of larger-scale folds in fold belts, when, as discussed in 2.4, theory and
shear strain across lithological boundaries, thus yielding informa- modeling have shown that multilayers comprising numerous layers
tion on effective viscosity ratios (which are the inverse of the shear- will fold with a stronger amplification than a single or a few layers.
strain ratio). Examples are illustrated in Treagus (1999, Fig. 4) for This might suggest that large folds affecting numerous layers would
using cleavage refraction on fold limbs (e.g. Fig. 31a,b) to estimate fold more strongly than smaller folds affecting one or two layers,
approximate viscosity ratios of greywacke/slate. A gradual change whereas field studies and minor-major fold relationships suggest
of cleavage orientation in a graded bed (e.g. Fig. 31b) can also the opposite. Treagus and Fletcher (2009) find that small folds in
indicate ‘way up’, a feature that can be useful in mapping of one layer are likely to outgrow larger multilayered folds, with the
geological structures. In a related approach, Treagus et al. (2003) potential to become minor (parasitic) folds, if the thin layer is more
illustrated how the sense of cleavage refraction around a much- competent than the rest, and/or if the multilayer is narrowly or
studied anticline in Britain helps to reveal its two-phase fold stiffly confined. Fig. 15 illustrates the process, and the progressive
history and hinge migration. change from initially symmetric to asymmetric minor (parasitic)
folds around a major fold with progressive deformation, as classi-
6.2. Information from asymmetry of folds on different scales cally proposed by Ramberg (1964). A natural example is shown in
Fig. 32a, and strongly asymmetric folds that may lie on the flank of
Folds are seen on a variety of scales in orogenic belts, and a long- a larger fold in Fig. 32b. Other examples of folds on two or more
held method used by structural geologists in the field is to use the different scales are shown in Figs. 1d and 2a.
asymmetry and vergence of small-scale folds to indicate the Ghosh and Sengupta (2010) describe a somewhat different
geometry of larger-scale folds. In early to mid-20th Century struc- situation in which multilayer folds in mica schist sandwiched
tural geology texts, small-scale asymmetric folds were termed drag between more competent quartzite layers display the asymmetry
folds (e.g. Leith, 1923, p. 176; Billings, 1954, p. 78) and envisaged to of parasitic folds around the longer wavelength folds in the
be the result of shearing or flexural slip on larger-scale fold limbs, quartzite. It is the anisotropy of the mica schist in this case that
and were also termed parasitic folds (de Sitter, 1964, p. 279; Ramsay, results in internal folding (see Section 2.4), which has a greater
1967, p. 396). However, seminal papers by Ramberg (1963, 1964) buckling instability than the buckling of the quartzite layers in an
established that asymmetric minor folds would have initiated as anisotropic matrix.

Fig. 32. (a) Parasitic folds in quartz vein in schist changing symmetry around a larger structure, Archean rocks near Geraldton, Ontario; coin 2.4 cm. (b) Strongly asymmetric buckle
folds with thickened hinges and cleavage at a low angle to average bedding, in limestone layer in slates, Hele Bay, Devon, England; compass 15 cm. (c) Cross-section of sheath fold in
quartzite in Caledonide cover nappes, Trollheimen, Norway; knife 9 cm. See Hansen (1971, Fig. 34) for a 3D representation of this fold pattern. (d) Asymmetric folds in quartz vein in
iron-rich schist, reflecting the sense of displacement in a dextral shear zone, Archean rocks near Geraldton, Ontario; coin 1.9 cm.
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2065

Minor folds not only provide valuable information from their ‘S’ Holdsworth, 2002). Carreras et al. (2005) provide a good review
and ‘Z’ asymmetries (fold vergence; Bell, 1981) that reveal larger- of the kinematics of folds of various types in shear zones.
scale and regional folds and aid geological mapping. Results in Mechanical instabilities can lead to folding in layered or aniso-
Treagus and Fletcher (2009) suggest that the relative scales of tropic rocks in shear zones. Buckling instabilities in single layers
folding, and the presence of smaller with larger folds, can poten- will exist if the layer is oblique to the shear zone boundary and
tially provide information on the relationships of rheology and there is a component of shortening parallel to the layer. Initially the
stratigraphy in multilayered rocks. This work ties together two folds produced are symmetric, and they become modestly asym-
classical studies and concepts: the structural lithic units of Currie metric with increasing strain (Ghosh, 1966; Manz and Wickham,
et al. (1962); and the orders of folding of Ramberg (1964). 1978). Asymmetry of cleavage refraction in the two limbs is more
Most of this review has concentrated on the theory and marked than the asymmetry of fold shape, with stronger refraction
modeling of buckle folds that develop in single or multiple layers in the long limb than in the short limb (Viola and Mancktelow,
that are parallel to the principal shortening deformation. The usual 2005). In multilayered, effectively anisotropic media or homoge-
assumption is horizontal layering, horizontal shortening, and neous anisotropic media, folding instabilities exist, the effects of
symmetric upright folds. However, folds in nature, even those that which depend on the orientation of the plane of anisotropy to the
result from one phase of deformation, are rarely perfectly shear plane. The structures produced include kink bands, crenula-
symmetrical. Large-scale fold asymmetry, and inclined fold axial tions and shear bands. They have been studied experimentally
planes, are common field observations and can have many expla- (Ramberg and Johnson, 1976; Williams and Price, 1990; Mandal
nations, some of which are suggested here. Local and variable et al., 2004) and treated conceptually and analytically (Platt,
asymmetry of buckle folds will be inherited from the initial 1983; Dennis and Secor, 1987; Fletcher, 2005). Fletcher (2005)
perturbation spectrum, as seen in analog and numerical models showed that a weak instability existed even for the case of the
(e.g. Figs. 10 and 14, Mancktelow, 2001). Perhaps the simplest plane of anisotropy parallel to the shear plane, although amplifi-
reason for systematic fold asymmetry in the fold profile view, cation is followed by deamplification as the structure rotates under
investigated by the second author many years ago (Beech, 1969; the influence of the shear.
Treagus, 1972), is that the layers are oblique to the principal Flanking structures are isolated folds and shear bands associated
shortening, but parallel to the intermediate strain axis (Y). Layering with strain perturbation around veins, dikes or fractures in shear
is contracted but also undergoes rotation and shearing, progres- zones (Fig. 33a, Hudleston, 1989; Passchier, 2001; Grasemann and
sively changing folds that initiate as symmetric waves (Treagus, Stüwe, 2001). Since different mechanisms can produce these
1973), into asymmetric folds (Anthony and Wickham, 1978). The structures, care must be used when using them as sense of shear
fold axis would be expected to parallel Y, but the axial plane would indicators (Passchier, 2001; Grasemann and Stüwe, 2001). Kocher
initiate perpendicular to layering, and only approach the bulk XY and Mancktelow (2006) show that anisotropy significantly influ-
plane when the folds are quite tight and asymmetric. Such ences the development of flanking structures, which must be taken
a scenario might be initially dipping stratigraphy, such as on into account when considering these structures as shear sense
a continental shelf, undergoing subhorizontal crustal shortening. indicators. They also suggest that the geometry of these structures
Buckle folds in a wedge-shaped fold-thrust belt would develop may provide an estimate of the degree of anisotropy of rocks in
with a polarity and regional sense of overturning or vergence that shear zones.
reflects the asymmetry of and shear within the wedge (e.g. Rowan
and Kligfield, 1992), and fault-related folds in fold-thrust belts are 6.4. Three-dimensional features of folds: localization, transected
asymmetric reflecting the geometry and sequence of development folds, and refolded folds
of the thrusts with which they are associated. (e.g. Suppe, 1983,
1985) Until now, we have mainly concentrated on folds in single or
Another major cause of fold asymmetry, which has important multiple layers, in theory, models and naturally-deformed rocks,
regional structural and tectonic implications, is folding associated and with the assumption that the folds form by processes of
with shear zones, discussed in the following section. buckling. The folds initiate as a series of sinusoidal-type waves that
are explicable accordingly to buckling theory and modeling for the
6.3. Folds in shear zones rheology concerned, as discussed in section 2. We would argue that
this approach is valid for a wide range of geological settings for
Folds are commonly found in ductile shear zones, which are ductile rocks in orogenic belts. However, a geologist studying
common in the mid to deep levels of the crust and in weak mate- higher-level sedimentary rocks, upon discovering one or more
rials such as evaporites at higher levels. The strain gradients that folds, might come to a different conclusion. For him or her, the fold
define ductile shear zones may produce passive folds in layering might be interpreted as an important localized structure, indicating
that lacks significant rheological contrasts, and similar folds with faulting at depth or out of section. Buckling might not be regarded
axial surfaces sub-parallel to the shear zone boundaries reflect the as the primary cause, unless a sequence of sinusoidal folds are seen.
combination of large shear strains with irregularities in the layering Folds that occur as localized features, whether they affect soft
and low rheological contrasts (e.g. Fig. 31d, Hudleston, 1977; sediments (e.g. Woodcock, 1976), are forced or drape folds (e.g.
Carreras et al., 2005). Fold hinges typically form at a high angle to Chester et al., 1988; Cook, 1988), or are variously fault-related
the shear direction, but with progressive strain, they tend to rotate (Suppe, 1983, 1985; McClay, 1992), provide important information
towards parallelism with the shear direction (e.g. Escher and about the conditions of the rock and the local deformation, but the
Watterson, 1974), in three dimensions forming sheath folds (e.g. varieties are too great to contain in this review. For fault-related
Cobbold and Quinquis, 1980; Skjernaa, 1989; Fig. 32c), which are folds, readers are referred to a Special Issue of this Journal
a key indicator of sense of shear in shear zones (e.g. Simpson, and (Wilkerson et al., 2002) for information and review.
Schmid, 1983; Hanmer and Passchier, 1991). Fold asymmetry is Another example of localized folding would be the process of
also used as a shear-sense indicator (e.g. Fig. 32d), but this must be folding according to the thermal-mechanical feedback model of
applied with care, as reversals in apparent sense of shear can arise Hobbs et al. (2008), by which thermally induced weakening
in a number of ways, as a result of large strains and 3D effects promotes buckling at low values of l/h, with deformation concen-
(Ramsay et al., 1983; Passchier and Williams, 1996; Alsop and trated in shear zones in the stiff layer connecting adjacent inner arc
2066 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

Fig. 33. (a) Flanking folds associated with rotated fractures/veins in a dextral shear zone in Archean biotite schist, Rainy Lake, Minnesota; coin 2.4 cm. White arrows indicate
fractures about which flanking folds formed. (b) Cleavage-transected fold in sandstones and mudstones, Southern Uplands, Scotland (Stringer and Treagus, 1980, Fig. 5c); hammer
head 18 cm. The hammer, parallel to the fold axis, is at a clear angle to the axial-planar cleavage. (c) F1/F2 fold interference marked by passive behavior of layers in Moine schists.
Loch Monar, Scotland; coin 3 cm. Axial surface traces of 2 F2 folds marked by dashed lines. (d) F1/F2 fold interference complicated by buckling of white pegmatitic veins in both F1
and F2 deformations, Maggia nappe, Ticino, Switzerland; coin 2.3 cm.

fold hinges. We and others have recently discussed and questioned recognised to contain components of transpression (Harland, 1971;
the validity of this model as a significant process of small-scale Sanderson and Marchini, 1984), has led to conclusions that clock-
geological folding (Treagus and Hudleston, 2009; Hobbs et al., wise-transected folds are direct evidence for sinistral transpressive
2009, 2010; Schmid et al., 2010), and will not repeat the argu- deformation (Soper and Hutton, 1984; Soper et al., 1987; Woodcock
ments here. et al., 1988), and vice versa. However, Treagus and Treagus (1992)
Folds in which the cleavage is observed to be oblique to the fold questioned the theoretical basis for equating transected folds
axis and axial plane, i.e. non-axial-planar, are termed transected with transpression, and concluded that there is no obvious expla-
folds (Borradaile, 1978; Stringer and Treagus, 1980). An example is nation for transected folds in horizontal layers in simple trans-
shown in Fig. 33b. One explanation is buckle folding of layers that pression (Harland, 1971; Sanderson and Marchini, 1984), unless
are oblique to all three principal directions of strain (Flinn, 1962; cleavage does not does not track the finite XY plane of strain (see
Treagus and Treagus, 1981): the fold axis for a 3D-oblique layer Soper, 1986). For the case where layers have different initial
(parallel to the longer axis of the sectional ellipse) would not orientations (strikes and dips) with respect to a horizontal trans-
generally coincide with the intersection of the XY principal plane pression, analogous to folding of oblique layers discussed above,
and the layer. For this type of oblique deformation, assuming transected folds would now be expected, but there is no simple
cleavage forms parallel to the XY plane, the cleavage would be relationship between, for example, clockwise transection and
oblique to the axial plane. Treagus and Treagus (1981) investigated sinistral transpression: the relationships would vary according to
the structures that would be associated with folding of generally the orientations of the folding layers (see Treagus and Treagus,
oblique layers in a triaxial strain ellipsoid, showing that transection 1992, for more examples and values).
angles (Borradaile, 1978) rarely exceed 10 , but increase from flat- We now turn to polyphase deformation, and the information
tening to constrictional strain (i.e. with the strain ellipsoid shape that polyphase and refolded folds (e.g. Fig. 33c,d) can provide about
factor; Flinn, 1965). As significant as the transection are the the deformation of an area. The patterns of refolding and fold
expected values for axial (hinge) migration, fold plunge variations, interference, classically described by O’Driscoll (1962a, 1962b) and
and the degree of cylindroidal or periclinal geometry, for layers Ramsay (1962b, 1967, chapter 10) and classified by Thiessen and
folding in constrictional to flattening strain ellipsoids (Treagus and Means (1980) and Grasemann et al. (2004), make up one of the
Treagus, 1981). Periclinal folding, strong plunge variations, and modeling sessions in Ramsay and Lisle (2000, session 35). The two
folding in two directions (cross folding) were suggested as indica- interfering fold phases are both modeled as sinusoidal fold-forming
tors of folding in a constrictional deformation environment. simple shear deformations (i.e. similar folds): the ‘interference’
The occurrence of transected folds in the Caledonian-Appala- map patterns are produced geometrically, without the involvement
chian orogenic belt (e.g. Fig. 33b), where the deformation is of mechanics of buckling. This is a useful way of determining the
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2067

characteristic geometry of fold interference (see Ramsay, 1967, write about one of our favorite topics, a happy reminder both of our
Fig. 10e13) according to the relative orientations of the two fold research begun long ago at Imperial College, London, and of our
phases: patterns such as egg-box, crescent, etc. However, it is shared time as former editors of JSG. We would like to thank John
known from analog experimental studies in systems with rheo- Ramsay for inspiring us to appreciate the beauty of folds, and Ray
logical contrasts (Ghosh and Ramberg, 1968; Watkinson, 1981; Fletcher for providing keen comments, interest and the sharing of
Odonne and Vialon, 1987; Ghosh et al., 1992, 1993, 1995; Grujic, ideas. PJH acknowledges earlier support from the National Science
1993; Johns and Mosher, 1996; Sengupta et al., 2005), that the Foundation and support in various forms from the University of
geometry of the ‘first folds’ will influence the ‘second folds’. The Minnesota. SHT acknowledges support for research from the NERC
studies by Ghosh and others reveal the 3D complexities that arise in as a Senior Research Fellow and from the University of Manchester,
interfering folds on different scales in materials with rheological and thanks Jack Treagus for his constant interest and encourage-
contrasts. The folds shown in Fig. 33d involved buckling during two ment over the years. Finally, we thank Neil Mancktelow and Sudipta
separate phases of deformation, this being mechanically easiest Sengupta for their review comments that have been invaluable in
when, as here, folds of the two phases are coaxial. helping us revise the paper.
Watkinson and Cobbold (1981) showed that earlier fold axes or
lineations will act as a linear anisotropy, and may ‘control’ the Appendix. Symbols used
orientation of later fold axes. Hence, it should not necessarily be
assumed that the axis of later folding and small-scale fold hinges
and crenulations, exactly reflect a position perpendicular to the A Amplitude
maximum shortening direction for the later deformation. Never- Ao Initial amplitude
theless, and despite these reservations, the patterns of refolding, B Constant in power law
the presence of small-scale folds and crenulated fabrics, and the e_ Strain rate
orientations of fold axes and fabrics of different generations EL/EM Elasticity (Young’s modulus) ratio of layer to matrix
provide a valuable resource for gaining information about regional G Elastic shear modulus
deformations and tectonics, and have played a historically impor- h Thickness of individual layers
tant part in documenting and unravelling the deformation of many ho Initial thickness of individual layers
parts of the Earth’s crust (see Turner and Weiss, 1963 and Ramsay, H Thickness of multilayer package
1967 for seminal examples of earlier work). k Wave number, k ¼ 2ph=l
ko Initial wavenumber
7. Future work Ki Curvature index, defined in Fig. 18.
L Arclength
The mechanical processes that produce folds in rocks are now Lo Initial arclength
fairly well understood, based on a solid body of theoretical and L=h Mean value of L/h for a fold population
modeling work, tied to a number of careful field studies. Questions n Power law exponent
remain, however. In the past dozen years or so, attention has been n* Effective power-law exponent in non-plane strain
paid to possible elastic effects, in addition to the clear importance of conditions
viscous behavior, on fold development. If rocks behave in a visco- nL, nM Power law exponents of layer and matrix
elastic manner (with Maxwell or more complex rheology) stress N Number of stiff layers in a multilayer package, or total
level or strain rate becomes important in determining how folds layers in a package
develop - if the strain rate or stress level is high enough - and non- P Layer-parallel compressive stress
linear effects may become significant. It is not yet clear, however, if q(k) Growth factor
pffiffiffiffiffi
elastic properties are important in natural fold development, Q Rheological parameter in Eq. (8), Q ¼ nL R00
especially for the small-scale folds so common in most orogenic R Ratio of viscous to elastic dominant wavelengths,
belts. It will be important to develop criteria that allow determi- R ¼ ldv/lde
nation of whether or not elastic properties play a role in natural fold R0 Ratio of normal viscosities for anisotropic buckling,
development. Some of the work reviewed here does suggest that R0 ¼ m0nM =mL
0n

such criteria may be developed. Other processes or phenomena R00 Effective viscosity
qffiffiffiffiffiffiffiffiffiffiffiffiffi ratio for anisotropic buckling,
that have been suggested as controlling fold development, R00 ¼ m0n M M
m 0s =m0n
L
including thermal-mechanical feedback, also need to be tested s Spacing of competent layers in a multilayer package
against well-constrained field examples. S Stretch
The greatly increased power of modern computers allows S1, S2 Principal stretches in plane strain (the xz plane), with
increasingly realistic deformational scenarios to be modeled, and S1  S2 and S1S2 ¼ 1
this offers the prospect that both forward and reverse modeling Sc Crossover stretch
may be carried out in efforts to find the combination of rock Ss Scaled stretch
properties and deformational conditions that best accounts for t Time
observed structures. With increasingly sophisticated tools to t0 Normalized thickness in Ramsay thickness-dip plot
determine conditions of temperature, pressure and measurement T Ratio of principal stretches, T ¼ S1/S2
of age, and thus strain rate, and increased ability to interpret x,y,z Coordinate system, with x parallel to the shortening
microstructures in terms of rheological properties, strain rate and direction in plane strain, z perpendicular to x in the plane
state of strain, the possibility exists of tightly constraining condi- of strain.
tions of deformation and mechanical controls on natural folds. q Limb dip
a1, a2 Fractions of thicknesses of layers of viscosities m1 and m2 in
Acknowledgments multilayer package
aL, bL Functions of nL, defined in Eq. (9)
We thank Cees Passchier for inviting us to contribute a review d Anisotropy factor, d ¼ mn/ms
paper to mark JSG’s 30th anniversary. It brought us together to dL,dM Anisotropy factors of layer and matrix
2068 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

l Wavelength Casey, M., Huggenberger, P., 1985. Numerical modelling of finite-amplitude similar
ld Dominant wavelength folds developing under general deformation histories. Journal of Structural
Geology 7, 103e114.
lp Preferred wavelength Chamberlin, R.T., 1910. The Appalachian folds of central Pennsylvania. Journal of
lo Initial wavelength Geology 18, 228e251.
m Viscosity Chapple, W.M., 1968. A mathematical theory of finite amplitude rock-folding.
Geological Society of America Bulletin 79, 47e68.
mL,mM Viscosities of layer and matrix Chester, J.C., Spang, J.H., Logan, J.M., 1988. Comparison of thrust fault rock models to
mn, ms Normal and shear viscosities basement-cored folds in the Rocky Mountain foreland. In: Schmidt, C.J.,
mnL ,msL Normal and shear viscosities of layer (and similar for Perry Jr., W.J. (Eds.), Interactions of the Rocky Mountains Foreland and the
Cordilleran Thrust Belt. Geological Society of America Memoir, vol.171, pp. 65e74.
matrix) Claypole, E.W., 1885. Pennsylvania before and after the elevation of the Appa-
m0 L and m0 M Viscosities under the base rate of flow lachian mountains, a study in dynamical geology. The American Naturalist
n Poisson’s ratio 19, 257e268.
Cloetingh, S., Burov, E., Beekman, F., Andeweg, B., Andriessen, P.A.M., Garcia-
x Ratio of strain rates in x and y directions, e_ y =e_ x , in buckling Castellanos, D., de Vicente, G., Vegas, R., 2002. Lithospheric folding in Iberia.
under non-plane strain Tectonics 21 (5), 26.
sx Total normal stress in layer Cobbold, P.R., 1975. Fold propagation in single embedded layers. Tectonophysics 27,
333e351.
sx Uniform normal stress in layer (membrane stress) Cobbold, P.R., 1976a. Mechanical effects of anisotropy during large finite deforma-
s~x Normal fiber stress in layer tion. Bulletin de la Société géologique de France 18, 1497e1510.
s Dimensionless time, s ¼ e_ x t, Cobbold, P.R., 1976b. Fold shapes as functions of progressive strain. Philosophical
4 Maximum stretch direction in fold profile Transactions of the Royal Society A 283, 129e138.
Cobbold, P.R., 1977. The finite element analyses of fold propagation e problematic
application. Tectonophysics 38, 339e353.
Cobbold, P.R., Quinquis, H., 1980. Development of sheath folds in shear regimes:
References Shear zones in rocks. Journal of Structural Geology 2, 119e126.
Cobbold, P.R., Cosgrove, J.W., Summers, J.M., 1971. Development of internal struc-
Abbassi, M.R., Mancktelow, N.S., 1990. The effect of initial perturbation shape and tures in deformed anisotropic rocks. Tectonophysics 12, 23e53.
symmetry on fold development. Journal of Structural Geology 12, 273e282. Cook, D.G., 1988. Balancing basement-cored folds of the Rocky Mountains foreland.
Aller, J., Bastida, F., Lisle, R.J., Ramsay, J.G., 2010. Photograph of the month: flexural In: Schmidt, C.J., Perry Jr., W.J. (Eds.), Interactions of the Rocky Mountains
slip folding of foresets in cross-bedded sandstones. Journal of Structural Foreland and the Cordilleran Thrust Belt. Geological Society of America Memoir,
Geology, 725e726. vol. 171, pp. 53e64.
Alsop, G.I., Holdsworth, R.E., 2002. The geometry and kinematics of flow pertur- Cruikshank, K.M., Johnson, A.M., 1993. High amplitude folding of linear-viscous
bation folds. Tectonophysics 350, 99e125. multilayers. Journal of Structural Geology 15, 79e94.
Anthony, J.M., Wickham, J.S., 1978. Finite-element simulation of asymmetric folding. Currie, J.B., Patnode, H.W., Trump, R.P., 1962. Development of folds in sedimentary
Tectonophysics 47, 1e14. strata. Geological Society of America Bulletin 73, 655e674.
Bastida, F., Aller, J., Bobillo-Ares, N.C., 1999. Geometrical analysis of folded surfaces Czeck, D.M., Fissler, D.A., Horsman, E., Tikoff, B., 2009. Strain analysis and rheology
using simple functions. Journal of Structural Geology 21, 729e742. contrasts in polymictic conglomerates: an example from the Seine meta-
Bastida, F., Aller, J., Bobillo-Ares, N.C., Toimil, N.C., 2005. Fold geometry; a basis for conglomerates, Superior Province, Canada. Journal of Structural Geology 31,
their kinematical analysis. Earth-Science Reviews 70, 129e164. 1365e1376.
Bastida, F., Aller, J., Toimil, N.C., Lisle, R.J., Bobillo-Ares, N.C., 2007. Some considerations Dahlstrom, C.D., 1969. Balanced cross sections. Canadian Journal of Earth Sciences 6,
on the kinematics of chevron folds. Journal of Structural Geology 29, 1185e1200. 743e757.
Bayly, M.B., 1970. Viscosity and anisotropy estimates from measurements on de Sitter, L.U., 1964. Structural Geology. McGraw Hill, New York.
chevron folds. Tectonophysics 9, 459e474. Dennis, A.J., Secor, D.T., 1987. A model for the development of crenulations in shear
Bayly, M.B., 1974. An energy calculation concerning the roundness of folds. Tecto- zones with applications from the Southern Appalachian Piedmont. Journal of
nophysics 24, 291e316. Structural Geology 9, 809e817.
Beech, S.H., 1969. Buckling of single and multi-layers oblique to principal Dieterich, J.H., 1969. Origin of cleavage in folded rocks. American Journal of Science
compressive stress, and its bearing on the problem of asymmetric fold forma- 267, 155e165.
tion. [Link]. thesis (unpublished), Imperial College, University of London. Dieterich, J.H., Carter, N.L., 1969. Stress history of folding. American Journal of
Bell, A.M., 1981. Vergence: an evaluation. Journal of Structural Geology 3, 197e202. Science 267, 129e154.
Billings, M.P., 1954. Structural Geology. Prentice-Hall Inc., New Jersey. Escher, A., Watterson, J., 1974. Stretching fabrics, folds and crustal shortening.
Biot, M.A.,1961. Theory of folding of stratified viscoelastic media and its implications in Tectonophysics 22, 223e231.
tectonics and orogenesis. Geological Society of America Bulletin 72, 1595e1620. Fletcher, R.C., 1974. Wavelength selection in the folding of a single layer with
Biot, M.A., 1964. Theory of internal buckling of a confined multilayered structure. power-law rheology. American Journal of Science 274, 1029e1043.
Geological Society of America Bulletin 75, 563e568. Fletcher, R.C., 1977. Folding of a single viscous layer: exact infinitesimal-amplitude
Biot, M.A., 1965a. Mechanics of Incremental Deformations. John Wiley and Sons, solution. Tectonophysics 39, 593e606.
New York. Fletcher, R.C., 1979. The shape of single-layer folds at small but finite amplitude.
Biot, M.A., 1965b. Theory of similar folding of the first and second kind. Geological Tectonophysics 60, 77e87.
Society of America Bulletin 76, 251e258. Fletcher, R.C., 1991. 3-dimensional folding of an embedded viscous layer in pure
Biot, M.A., 1965c. Further development in the theory of internal buckling of shear. Tectonophysics 13, 87e96.
multilayers. Geological Society of America Bulletin 76, 833e840. Fletcher, R.C., 1995. 3-dimensional folding and necking of a power-law layer e are
Biot, M.A., Ode, H., Roever, W.L., 1961. Experimental verification of the theory of folds cylindrical, and if so, do we understand why? Tectonophysics 247, 65e83.
folding of stratified viscoelastic media. Geological Society of America Bulletin Fletcher, R.C., 2005. Instability of an anisotropic power-law fluid in a basic state of
72, 1621e1632. plane flow. Journal of Structural Geology 27, 1155e1167.
Bobillo-Ares, N.C., Bastida, F., Aller, J., 2000. On tangential longitudinal strain Fletcher, R.C., Sherwin, J., 1978. Arc lengths of single layer folds: a discussion of the
folding. Tectonophysics 319, 53e68. comparison between theory and observation. American Journal of Science 278,
Bobillo-Ares, N.C., Toimil, N.C., Aller, J., Bastida, F., 2004. FoldModeler; a tool for the 1085e1098.
geometrical and kinematical analysis of folds. Computers and Geosciences 30, Flinn, D., 1962. On folding during three-dimensional progressive deformation.
147e159. Quarterly Journal Geological Society London 118, 385e433.
Bobillo-Ares, N.C., Aller, J., Bastida, F., Lisle, R.J., Toimil, N.C., 2006. The problem of Flinn, D., 1965. Deformation in metamorphism. In: Pitcher, W.S., Flinn, G.W. (Eds.),
area change in tangential longitudinal strain folding. Journal of Structural Controls of Metamorphism. Geological Journal Special Issue No 1. Oliver and
Geology 28, 1835e1848. Boyd, Edinburgh and London, pp. 48e72.
Bobillo-Ares, N.C., Bastida, F., Aller, J., Lisle, R.J., 2009. An approach to folding Frehner, M., Schmalholz, S.M., 2006. Numerical simulations of parasitic folding in
kinematics from the analysis of folded oblique surfaces. Journal of Structural multilayers. Journal of Structural Geology 28, 1647e1657.
Geology 31, 841e852. Gairola, V.K., Kern, H., 1984. Single-layer folding in marble and limestone: an
Borradaile, G.J., 1978. Transected folds: a study illustrated with examples from experimental study. Tectonophysics 108, 155e172.
Canada and Scotland. Geological Society of America Bulletin 89, 481e493. Ghosh, S.K., 1966. Experimental tests of buckling folds in relation to strain ellipsoid
Burg, J.-P., Podladchikov, Y., 1999. Lithospheric scale folding: numerical modelling in simple shear deformations. Tectonophysics 3, 169e185.
and application to the Himalayan syntaxes. International Journal of Earth Ghosh, S.K., 1968. Experiments on buckling of multilayers which permit interlayer
Sciences 88, 190e200. gliding. Tectonophysics 6, 207e249.
Carreras, J., Druguet, E., Griera, A., 2005. Shear zone-related folds. Journal of Ghosh, S.K., 1993. Structural Geology: Fundamentals and Modern Developments.
Structural Geology 27, 1229e1251. Pergamon, Oxford.
Carter, N.L., Tsenn, M.C., 1987. Flow properties of continental lithosphere. Tecto- Ghosh, S.K., Ramberg, H., 1968. Buckling experiments on intersecting fold patterns.
nophysics 136, 27e63. Tectonophysics 5, 89e105.
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2069

Ghosh, S.K., Sengupta, S., 2010. Paradoxical situation in determining relative Johns, M.K., Mosher, S., 1996. Physical models of regional fold superposition: the
competence from wavelength/arclength ratios within buckle folded multi- role of competence contrast. Journal of Structural Geology 18, 475e492.
layers. Journal of the Geological Society of India 75, 13e17. Johnson, A.M., 1970. Physical Processes in Geology. Freeman, Cooper and Company,
Ghosh, S.K., Mandal, N., Khan, D., Deb, S.K., 1992. Modes of superposed buckling in San Francisco.
single layers controlled by initial tightness of early folds. Journal of Structural Johnson, A.M., 1977. Styles of Folding. Elsevier, New York.
Geology 14, 381e394. Johnson, A.M., Fletcher, R.C., 1994. Folding of Viscous Layers. Columbia University
Ghosh, S.K., Mandal, N., Sengupta, S., Deb, S.K., Khan, D., 1993. Superposed buckling Press, New York.
in multilayers. Journal of Structural Geology 15, 95e111. Johnson, A.M., Pfaff, V.J., 1989. Parallel, similar and constrained folds. Engineering
Ghosh, S.K., Khan, D., Sengupta, S., 1995. Interfering folds in constrictional defor- Geology 27, 115e180.
mation. Journal of Structural Geology 17, 1361e1373. Kaus, B.J.P., Schmalholz, S.M., 2006. 3D finite amplitude folding; implications for
Grasemann, B., Stüwe, K., 2001. The development of flanking folds during simple shear stress during crustal and lithospheric deformation. Geophysical Research
and their use as kinematic indicators. Journal of Structural Geology 23, 715e724. Letters 33, L14309.
Grasemann, B., Wiesmayr, G., Dragantis, E., Fusseis, F., 2004. Classification of refold Kirby, S.H., Kronenberg, A.K., 1987. Rheology of the lithosphere; selected topics.
structures. Journal of Geology 112, 119e125. Reviews of Geophysics 25, 1219e1244.
Gray, D.R., 1981. Cleavageefold relationships and their implications for transected Kocher, T., Mancktelow, N.S., 2006. Flanking structure development in anisotropic
folds: an example from southwest Virginia, U.S.A. Journal of Structural Geology viscous rock. Journal of Structural Geology 28, 1139e1145.
3, 265e277. Kocher, T., Schmalholz, S.M., Mancktelow, N.S., 2006. Impact of mechanical anisotropy
Grujic, D., 1993. The influence of initial fold geometry on Type 1 and Type 2 and power-law rheology on single layer folding. Tectonophysics 421, 71e87.
interference patterns: an experimental approach. Journal of Structural Geology Kocher, T., Mancktelow, N.S., Schmalholz, S.M., 2008. Numerical modelling of the
15, 293e307. effect of matrix anisotropy orientation on single layer fold development. Journal
Grujic, D., Mancktelow, N.S., 1997. Folds with axes parallel to the extension direc- of Structural Geology 30, 1013e1023.
tion; an experimental study. Journal of Structural Geology 17, 279e291. Kohlstedt, D.L., Evans, B., Mackwell, S.J., 1995. Strength of the lithosphere;
Hall, J., 1815. On the vertical position and convolutions of certain strata and their constraints imposed by laboratory experiments. Journal of Geophysical
relation with granite. Transactions Royal Society Edinburgh 7, 79e108. Research 100, 17,587e17,602.
Hanmer, S., Passchier, C., 1991. Shear-sense indicators; a review. Paper e Geological Lan, L., Hudleston, P.J., 1991. Finite-element models of buckle folds in non-linear
Survey of Canada 90-17, 72. materials. Tectonophysics 199, 1e12.
Hansen, E., 1971. Strain Facies. Springer-Verlag. New York. Lan, L., Hudleston, P.J., 1995a. A method of estimating the stress exponent in the
Harland, W.B., 1971. Tectonic transpression in Caledonian Spitsbergen. Geological flow law for rocks using fold shape. Pure and Applied Geophysics 145, 621e635.
Magazine 108, 27e42. Lan, L., Hudleston, P.J., 1995b. The effects of rheology on the strain distribution in
Hills, E.S., 1963. Elements of Structural Geology. Methuen, London. single layer buckle folds. Journal of Structural Geology 17, 727e738.
Hirth, G., Teyssier, C., Dunlap, W.J., 2001. An evaluation of quartzite flow laws based Lan, L., Hudleston, P., 1996. Rock rheology and sharpness of folds in single layers.
on comparisons between experimentally and naturally deformed rocks; Journal of Structural Geology 18, 925e931.
Deformation mechanisms, rheology and microstructures. Geologische Run- Latham, J.-P., 1979. Experimentally developed folds in a material with a planar
dschau e International Journal of Earth Sciences 90, 77e87. mineral fabric. Tectonophysics 57, T1eT8.
Hobbs, B., Regenauer-Lieb, K., Ord, A., 2008. Folding with thermal-mechanical Latham, J.-P., 1985a. The influence of nonlinear material properties and resistance to
feedback. Journal of Structural Geology 30, 1572e1592. bending on the development of internal structures. Journal of Structural
Hobbs, B., Regenauer-Lieb, K., Ord, A., 2009. Folding with thermal-mechanical Geology 7, 225e236.
feedback: a reply. Journal of Structural Geology 31, 752e755. Latham, J.-P., 1985b. A numerical investigation and geological discussion of the
Hobbs, B., Regenauer-Lieb, K., Ord, A., 2010. Folding with thermal-mechanical relationship between folding, kinking and faulting. Journal of Structural
feedback: another reply. Journal of Structural Geology 32, 131e134. Geology 7, 237e249.
Holst, T.B., 1987. Analysis of buckle folds from the early Proterozoic of Minnesota. Leith, C.K., 1923. Structural Geology. Henry Holt and Company, Inc., New York.
American Journal of Science 287, 612e634. Lisle, R., 1995. Geological Structures and Maps: a Practical Guide. Butterworth-
Hudleston, P.J., 1973a. An analysis of “single layer” folds developed experimentally Heinemann, Oxford.
in viscous media. Tectonophysics 16, 189e214. Lisle, R., 1997. A fold classification scheme based on a polar plot of inverse layer
Hudleston, P.J., 1973b. An analysis and interpretation of minor folds in the Moine thickness. In: Sengupta, S. (Ed.), Evolution of Geological Structures in Micro to
rocks of Monar, Scotland. Tectonophysics 16, 89e132. Macro-scales. Chapman & Hall, London, pp. 323e339.
Hudleston, P.J., 1973c. Fold morphology and some geometrical implications of Mancktelow, N.S., 1999. Finite-element modelling of single-layer folding in elasto-
theories of fold development. Tectonophysics 16, 1e46. viscous materials; the effect of initial perturbation geometry. Journal of Struc-
Hudleston, P.J., 1976. Recumbent folds in base of Barnes-Ice-Cap, Baffin Island, tural Geology 21, 161e177.
Northwest-Territories, Canada. Geological Society of America Bulletin 87, Mancktelow, N.S., 2001. Single-layer folds developed from initial random perturba-
1684e1692. tions; the effects of probability distribution, fractal dimension, phase, and ampli-
Hudleston, P.J., 1977. Similar folds, recumbent folds, and gravity tectonics in ice and tude. In: Koyi, H.A., Mancktelow, N.S. (Eds.), Tectonic Modeling: a Volume in Honor
rocks. Journal of Geology 85, 113e122. of Hans Ramberg. Geological Society of America Memoir, vol. 193, pp. 69e87.
Hudleston, P.J., 1986. Extracting information from folds in rocks. Journal of Mancktelow, N.S., Abbassi, M.R., 1992. Single layer buckle folding in non-linear
Geological Education 34, 237e245. materials; II, Comparison between theory and experiment. Journal of Structural
Hudleston, P.J., 1989. The association of folds and veins in shear zones. Journal of Geology 14, 105e120.
Structural Geology 11, 949e957. Mandal, N., Samanta, S.K., Chakraborty, C., 2004. Problem of folding in ductile shear
Hudleston, P.J., Holst, T.B., 1984. Strain analysis and fold shape in a limestone layer zones; a theoretical and experimental investigation. Journal of Structural
and implications for layer rheology. Tectonophysics 106, 321e347. Geology 26, 475e489.
Hudleston, P.J., Lan, L., 1993. Information from fold shapes. Journal of Structural Manz, R.E., Wickham, J., 1978. Experimental analysis of folding in simple shear.
Geology 15, 253e264. Tectonophysics 44, 79e90.
Hudleston, P.J., Lan, L., 1994. Rheological controls on the shapes of single-layer folds. McClay, K.R. (Ed.), 1992. Thrust Tectonics. Chapman Hall, London.
Journal of Structural Geology 16, 1007e1021. Mühlhaus, H.-B., Hobbs, B., Ord, A., 1994. The role of axial constraints on the
Hudleston, P.J., Stephansson, O., 1973. Layer shortening and fold shape development evolution of folds in single layers. In: Siriwardane, Zaman (Ed.), Computer
in the buckling of single layers. Tectonophysics 17, 299e321. Methods and Advances in Geomechanics. Balkema, Rotterdam, pp. 223e231.
Hudleston, P.J., Tabor, J.R., 1988. Strain and fabric development in a buckled calcite Mühlhaus, H.-B., Sakaguchi, H., Hobbs, B.E., 1998. Evolution of three-dimensional
vein and rheological implications. Bulletin of the Geological Institutions of the folds for a non-Newtonian plate in a viscous medium. Proceedings of the Royal
University of Uppsala 14, 79e94. Society London A 454, 3121e3143.
Hudleston, P.J., Treagus, S.H., Lan, L., 1996. Flexural flow folding: does it occur in Mühlhaus, H.-B., Moresi, L., Hobbs, B., Dufour, F., 2002. Large amplitude folding in finely
nature? Geology 24, 203e206. layered viscoelastic rock structures. Pure and Applied Geophysics 159, 2311e2333.
Hunt, G.W., Mühlhaus, H.-B., Whiting, A.I.M., 1997. Folding processes and solitary Neurath, C., Smith, R.B., 1982. The effect of material properties on growth rates of
waves in structural geology. Philosophical Transactions of the Royal Society folding and boudinage: experiments with wax models. Journal of Structural
London, A 355, 2197e2213. Geology 4, 215e229.
Hunt, G., Mühlhaus, H., Hobbs, B., Ord, A., 1996. Localized folding of viscoelastic Nevin, C.M., 1931. Principles of Structural Geology. Wiley, New York.
layers. Geologische Rundshau 85, 58e64. O’Driscoll, E.S., 1962a. Experimental patterns in superimposed similar folding.
Hunt, G.W., Edmunds, R., Budd, C.J., Cosgrove, J.W., 2006. Serial parallel folding with Journal of the Alberta Society of Petroleum Geologists 10, 145e167.
friction; a primitive model using cubic B-splines. Journal of Structural Geology O’Driscoll, E.S., 1962b. Models for superposed laminar flow folding. Nature 196,
28, 444e455. 1146e1148.
James, A.I., Watkinson, A.J., 1994. Initiation of folding and boudinage in wrench Odonne, F., Vialon, P., 1987. Hinge migration as a mechanism of superimposed
shear and transpression. Journal of Structural Geology 16, 883e893. folding. Journal of Structural Geology 9, 835e844.
Jeng, F.S., Huang, K.P., 2008. Buckling folds of a single layer embedded in matrix e Ormand, C.J., Hudleston, P.J., 2003. Strain paths of three small folds from the Appa-
theoretical solutions and characteristics. Journal of Structural Geology 30, lachian Valley and Ridge, Maryland. Journal of Structural Geology 25, 1841e1854.
633e648. Parrish, D.K., 1973. A nonlinear finite element fold model. American Journal of
Jeng, F.S., Lai, Y.C., Teng, M.H., 2002. Influence of strain rate on buckle folding of an Science 273, 318e334.
elasto-viscous single layer. Journal of Structural Geology 24, 501e516. Passchier, C.W., 2001. Flanking structures. Journal of Structural Geology 23, 951e962.
2070 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071

Passchier, C.W., Williams, P.R., 1996. Conflicting shear sense indicators in shear Skjernaa, L., 1989. Tubular folds and sheath folds; definitions and conceptual
zones; the problem of non-ideal sections. Journal of Structural Geology 18, models for their development, with examples from the Grapesvare area,
1281e1284. northern Sweden. Journal of Structural Geology 11, 689e703.
Paterson, M.S., Weiss, L.E., 1966. Experimental folding and deformation in phyllite. Smith, R.B., 1975. Unified theory of the onset of folding, boudinage, and mullion
Geological Society of America Bulletin 77, 343e374. structure. Geological Society of America Bulletin 86, 1601e1609.
Platt, J.P., 1983. Progressive refolding in ductile shear zones. Journal of Structural Smith, R.B., 1977. Formation of folds, boudinage, and mullions in non-Newtonian
Geology 5, 619e622. materials. Geological Society of America Bulletin 88, 312e320.
Price, N.J., Cosgrove, J.W., 1990. Analysis of Geological Structures. Cambridge Smith, R.B., 1979. The folding of a strongly non-Newtonian layer. American Journal
University Press, Great Britain. of Science 279, 272e287.
Ramberg, H., 1959. Evolution of ptygmatic folding. Norsk Geologisk Tiddskrift 39, Soper, N.J., 1986. Geometry of anastomosing solution cleavage in transpression
99e155. zones. Journal of Structural Geology 8, 937e940.
Ramberg, H., 1960. Relationships between length of arc and thickness of ptyg- Soper, N.J., Hutton, D.H.W., 1984. Late Caledonian sinistral displacements in Britain:
matically folded veins. American Journal of Science 258, 36e46. implications for a three-plate collision model. Tectonics 3, 781e794.
Ramberg, H., 1961. Contact strain and folding instability of a multilayered body Soper, N.J., Webb, B.C., Woodcock, N.H., 1987. Late Caledonian (Acadian) trans-
under compression. Geologische Rundshau 51, 405e439. pression in north-west England: timing, geometry, and geotectonic signifi-
Ramberg, H., 1963. Fluid dynamics of viscous buckling applicable to folding of cance. Proceedings of the Yorkshire Geological Society 46, 175e192.
layered rocks. Bulletin of the American Association of Petroleum Geologists 47, Srivastava, D.C., Lisle, R.J., 2004. Rapid analysis of fold shape using Bezier curves.
484e505. Journal of Structural Geology 26, 1553e1559.
Ramberg, H., 1964. Selective buckling of composite layers with contrasted rheo- Srivastava, D.C., Shah, J., 2006. A rapid method for strain estimation from flattened
logical properties, a theory for the simultaneous formation of several orders of parallel folds. Journal of Structural Geology 28, 1e8.
folds. Tectonophysics 1, 307e341. Srivastava, D.C., Shah, J., 2008. The `isogon rosette` method for rapid estimation of
Ramberg, H., 1970. Folding of laterally compressed multilayers in the field of gravity, strain in flattened folds. Journal of Structural Geology 30, 444e450.
I. Physics of the Earth and Planetary Interior 2, 203e232. Stabler, C.L., 1968. Simplified Fourier analysis of fold shapes. Tectonophysics 6,
Ramberg, H., 1981. Gravity, Deformation and the Earth’s Crust, Second ed. Academic 343e350.
Press, London. Stringer, P., Treagus, J.E., 1980. Non-axial planar S1 cleavage in the Hawick rocks of
Ramberg, I.B., Johnson, A.M., 1976. A theory of concentric, kink and sinusoidal the Galloway area, southern Uplands, Scotland. Journal of Structural Geology 2,
folding and of monoclinal flexuring of compressible, elastic multilayers; V, 317e331.
Asymmetric folding in interbedded chert and shale of the Franciscan Complex, Suppe, J., 1983. Geometry and kinematics of fault-bend folding. American Journal of
San Francisco Bay area, California. Tectonophysics 32, 295e320. Science 283, 684e721.
Ramsay, J.G., 1962a. The geometry and mechanism of formation of “similar” type Suppe, J., 1985. Principles of Structural Geology. Prentice Hall, New Jersey.
folds. Journal of Geology 70, 309e327. Talbot, C.J., 1979. Fold trains in a glacier of salt in southern Iran. Journal of Structural
Ramsay, J.G., 1962b. Interference patterns produced by the superposition of folds of Geology 1, 5e18.
similar type. Journal of Geology 70, 466e481. Talbot, C.J., Jackson, M.P.A., 1987. Internal kinematics of salt diapirs. Bulletin of the
Ramsay, J.G., 1967. Folding and Fracturing of Rocks. McGraw Hill, New York. American Association of Petroleum Geologists 71, 1068e1093.
Ramsay, J.G., 1974. Development of chevron folds. Geological Society of America Tanner, P.W.G., 1989. The flexural-slip mechanism. Journal of Structural Geology 11,
Bulletin 85, 1741e1754. 635e655.
Ramsay, J.G., Huber, M.I., 1987. The Techniques of Modern Structural Geology. In: Tentler, T., 2001. Experimental study of single layer folding in nonlinear materials.
Folds and Fractures, vol. II. Academic Press, London. In: Koyi, H.A., Mancktelow, N.S. (Eds.), Tectonic Modeling: a Volume in Honor of
Ramsay, J.G., Lisle, R.J., 2000. The Techniques of Modern Structural Geology. In: Hans Ramberg. Geological Society of America Memoir, vol. 193, pp. 89e99.
Applications of Continuum Mechanics in Structural Geology, vol. 3. Academic Thiessen, R.L., Means, W.D., 1980. Classification of fold interference patterns:
Press, London. a reexamination. Journal of Structural Geology 2, 311e316.
Ramsay, J.G., Casey, M., Kligfield, R., 1983. Role of shear in development of the Treagus, S.H., 1972. Processes in Fold Development. Ph.D. thesis, University of
Helvetic fold-thrust belt of Switzerland. Geology 11, 439e442. Manchester, Unpublished.
Roberts, D., Strömgård, K.-E., 1972. A comparison of natural and experimental strain Treagus, S.H., 1973. Buckling stability of a viscous single-layer system oblique to the
patterns around fold hinge zones. Tectonophysics 14, 105e120. principal compression. Tectonophysics 19, 271e289.
Rowan, M.G., Kligfield, R., 1992. Kinematics of large-scale asymmetric buckle folds Treagus, S.H., 1982. A new isogon e cleavage classification and its application to
in overthrust shear; an example from the Helvetic nappes. In: McClay, K.R. (Ed.), natural and model fold studies. Geological Journal 17, 49e64.
Thrust Tectonics. Chapman Hall, London, pp. 165e173. Treagus, S.H., 1983. A new theory of finite strain variation through contrasting layers,
Rutter, E.H., 1993. Experimental rock deformation; techniques, results and appli- and its bearing on cleavage refraction. Journal of Structural Geology 5, 351e358.
cations to tectonics. Geology Today 9, 61e65. Treagus, S.H., 1988. Strain refraction in layered systems. Journal of Structural
Sanderson, D.J., Marchini, W.R.D., 1984. Transpression. Journal of Structural Geology Geology 10, 517e527.
6, 449e458. Treagus, S.H., 1993. Flow variations in power-law multilayers: implications for
Schmalholz, S.M., 2006. Scaled amplification equation: a key to the folding history competence contrasts in rocks. Journal of Structural Geology 15, 423e434.
of buckled viscous single-layers. Tectonophysics 419, 41e53. Treagus, S.H., 1997. Modelling deformation partitioning in folds. In: Sengupta, S.
Schmalholz, S.M., Mancktelow, N.S., 2008. Estimation of palaeo-rheology from buckle (Ed.), Evolution of Geological Structures in Micro to Macro-scales. Chapman &
fold geometries. Bollettino Della Società Geologica Italiana 127, 227e230. Hall, London, pp. 341e372.
Schmalholz, S.M., Podladchikov, Y., 1999. Buckling versus folding: importance of Treagus, S.H., 1999. Are viscosity ratios of rocks measurable from cleavage refrac-
viscoelasticity. Geophysical Research Letters 26, 2641e2644. tion? Journal of Structural Geology 20, 895e901.
Schmalholz, S.M., Podladchikov, Y.Y., 2000. Finite amplitude folding: transition from Treagus, S.H., Fletcher, R.C., 2009. Controls of folding on different scales in multi-
exponential to layer length controlled growth. Earth and Planetary Science layered rocks. Journal of Structural Geology 31, 1340e1349.
Letters 181, 619e633. Treagus, S.H., Hudleston, P.J., 2009. Folding with thermal-mechanical feedback:
Schmalholz, S.M., Podladchikov, Y.Y., 2001. Strain and competence contrast esti- discussion. Journal of Structural Geology 31, 749e751.
mation from fold shape. Tectonophysics 340, 195e213. Treagus, J.E., Treagus, S.H., 1981. Folds and the strain ellipsoid; a general model.
Schmalholz, S.M., Podladchikov, Y.Y., Schmid, D.D., 2001. A spectral/finite difference Journal of Structural Geology 3, 1e17.
method for simulating large deformations of heterogeneous, viscoelastic Treagus, S.H., Treagus, J.E., 1992. Transected folds and transpression: how are they
materials. Geophysical Journal International 145, 199e208. associated? Journal of Structural Geology 14, 361e367.
Schmid, D.W., Podladchikov, Y.Y., 2006. Fold amplification rates and dominant Treagus, S.H., Treagus, J.E., 2002. Studies of strain and rheology of conglomerates.
wavelength selection in multilayer stacks. Philosophical Magazine 86, Journal of Structural Geology 24, 1541e1567.
3409e3423. Treagus, S.H., Treagus, J.E., Droop, G.T.R., 2003. Superimposed deformations and
Schmid, D.W., Dabrowski, M., Krotkiewski, M., 2008. Evolution of large amplitude their hybrid effects: the Rhoscolyn Anticline unravelled. Journal of the
3D fold patterns; a FEM study; Recent advances in computational geodynamics; Geological Society 160, 117e136.
theory, numerics and applications. Physics of the Earth and Planetary Interiors Turcotte, D.L., Schubert, G., 1982. Geodynamics. John Wiley & Sons Inc, New York.
171, 400e408. Turner, F.J., Weiss, L.E., 1963. Structural Analysis of Metamorphic Tectonites.
Schmid, D.W., Schmalholz, S.M., Mancktelow, N.S., Fletcher, R.C., 2010. Comment on McGraw-Hill, New York.
‘Folding with thermal-mechanical feedback’. Journal of Structural Geology 32, Twiss, R.J., 1988. Description and classification of folds in single surfaces. Journal of
127e130. Structural Geology 10, 607e623.
Sengupta, S., Ghosh, S.K., Deb, S.K., Khan, D., 2005. Opening and closing of folds in Van Hise, C.R., 1894. Principles of North American Pre-Cambrian geology. United
superposed deformation. Journal of Structural Geology 27, 1282e1299. States Geological Survey 16th Annual Report, 581e843.
Sherwin, J.-A., Chapple, W.M., 1968. Wavelengths of single layer folds: a comparison Viola, G., Mancktelow, N.S., 2005. From XY tracking to buckling; axial plane cleavage
between theory and observation. American Journal of Science 266, 167e179. fanning and folding during progressive deformation. Journal of Structural
Shimamoto, T., Hara, I., 1976. Geometry and strain distribution of single-layer folds. Geology 27, 409e417.
Tectonophysics 30, 1e34. Watkinson, A.J., 1975. Multilayer folds initiated in bulk plane strain, with the axis of
Simpson, C., Schmid, S.M., 1983. An evaluation of criteria to deduce the sense of no change perpendicular to the layering. Tectonophysics 28, T7eT11.
movement in sheared rocks. Geological Society of America Bulletin 94, Watkinson, A.J., 1981. Patterns of fold interference: influence of early fold shapes.
1281e1288. Journal of Structural Geology 3, 19e23.
 P.J. Hudleston, S.H. Treagus / Journal of Structural Geology 32 (2010) 2042e2071 2071

Watkinson, A.J., Cobbold, P.R., 1981. Axial directions of folds in rocks with linear/ Woodcock, N.H., 1976. Structural style in slump sheets; Ludlow series, Powys,
planar fabrics. Journal of Structural Geology 3, 211e217. Wales. Journal of the Geological Society of London 132, 399e415.
Weiss, L.E., 1972. The Minor Structures of Deformed Rocks. Springer, New York. Woodcock, N.H., Awan, M.A., Johnson, T.E., Mackie, A.H., Smith, R.D., 1988. Acadian
Wilkerson, M.S., Fischer, M.P., Apotria, T., 2002. Fault-related folds: the transition tectonics in Wales during Avalonia/Laurentia convergence. Tectonics 7,
from 2-D to 3-D. Special issue. Journal of Structural Geology 24 (4). 483e495.
Williams, P.F., 1979. The development of folds in a cross-laminated siltstone. Journal Zhang, Y., Hobbs, B.E., Ord, A.M., Mühlhaus, H.B., 1996. Computer simulation of
of Structural Geology 1, 19e30. single layer buckling. Journal of Structural Geology 18, 643e655.
Williams, P.F., Price, G.P., 1990. Origin of kinkbands and shear-band cleavage in Zhang, Y., Mancktelow, N.S., Hobbs, B.E., Ord, A.M., Mühlhaus, H.B., 2000. Numerical
shear zones; an experimental study. Journal of Structural Geology 12, 145e164. modelling of single-layer folding: clarification of an issue regarding the possible
Willis, B., 1891. Mechanics of the Appalachian structure. United States Geological effect of computer codes and the influence of initial irregularities. Journal of
Survey 13th Annual Report, 212e281. Structural Geology 22, 1511e1522.