Sie sind auf Seite 1von 174

ISBN No 81-901729-6-7

Structural Geology for Petroleum Geoscientists

Published by
Association of petroleum Geologists,
3rd Floor, Geology Division, S&T Building,
KDMIPE, ONGC,
9-Kaulagarh Road,
Dehradun 248001, India

Tel: +91-135-2795187, 2796565, 2758088


+91-22-24045330, +91-9869222409 (Mumbai Office)

Fax: +91-135-2758088
+91-22-24045330 (Mumbai Office)

www.apgindia.org
apg_india@rediffmail.com

Published: September 2006

All rights including the right to translate or to reproduce this book


or parts thereof except for brief otations are reserved

Cover Design and Layout: James Peters

Printed at

Allied Printers
Dehradun Ph. +91-135-2654505, 3290845

ii
Association of Petroleum Geologists Special Publication 3

Structural Geology for


Petroleum Geoscientists

Dilip K. Mukhopadhyay
IIT Roorkee

iii
To

My Mother and Teachers

iv
Preface
Structural geology is obviously one of the more important subjects for geoscientists working in
petroleum industry. Folds and faults in deformed rocks make traps for hydrocarbon accumulation. Also,
large-scale deformations, the so-called tectonics, control the architecture of petroliferous sedimentary
basins. It is the primary job of a structural geologist to interpret geological map and field data, and infer
geometry of large scale folds and faults. However, geoscientists with varied specializations and working
with different kinds of data may also be called upon to make structural interpretations. For example,
lineament maps prepared from air photographs or satellite based images are commonly interpreted in
terms of crustal-scale deformation or seismologists working with seismic reflection profiles routinely
interpret subsurface structural geometry. It is imperative that geoscientists with different specializations
working in oil industries have basic working knowledge on structural geology. A number of excellent
textbooks on structural geology are now available but it appears that many a petroleum geoscientists are
reluctant to pick up any of these books. This is probably due to the fact that the scopes of these books are
much wider than the requirements of petroleum geoscientists. In this publication, the focus is on topics
that I think should be of common interest to most petroleum geologists and geophysicists. I hope this will
be particularly useful to those who did not have a thorough grounding in structural geology during their
college/university days. Also, students in a bachelor level structural geology course may find this book
useful. Detailed discussion on all the topics covered in this publication can be found in any standard
textbook on structural geology. A list of such textbooks is given in the reference section, copies of which
are on my desk all the time.
I am grateful to Dr. James Peters, Secretary, Association of Petroleum Geologists for his continuous
encouragement. But for his persistent demand that the manuscript be completed within a fixed time frame,
I would have taken eternity to finalize the same. Dr. Premanand Mishra, Dr. R. Krishnamurti and Mrs.
Mamata Gupta are thanked for reading the manuscript cover to cover and locating innumerable mistakes.
However, I alone remain responsible for the mistakes that escaped scrutiny. I am also grateful to my
friends and colleagues at IIT Roorkee (formerly University of Roorkee) who maintain a congenial
academic and social environment where individuals can tread the path of their choice. But for the full
support of my family, this publication would not see the light of the day.

Roorkee Dilip K. Mukhopadhyay


05 September, 2006

v
Contents
Preface
1. Introduction 1-5
2. Planes and lines 7-14
3. Force and stress 15-23
4 Mohr circle 25-32
5. Strain 33-41
6. Stress-strain relation 43-48
7. Brittle fracture criteria 49-59
8. Faults: Morphology and classification 61-74
9. Thrust faults 75-86
10. Normal faults 87-97
11. Strike-slip faults 99-105
12. Folds: Geometry and classification 107-120
13. Fault-related folding 121-135
14. Balanced cross sections: Introduction 137-145
15. Section Construction 147-157
16. Section restoration 159-164
17. References 165-167

vi
1. Introduction

The rocks in many of the sedimentary basins with good prospect for hydrocarbon
reserves are highly deformed. The deformation is manifested by the presence of large-
scale faults and folds, some of which act not only as traps for hydrocarbon
accumulation but also control depositional systems of source, reservoir and seal facies.
Structural geology is the subject that deals with deformed rocks. It is one of the key
subjects in hydrocarbon exploration and research. A reliable interpretation of
subsurface structural geometry is essential for locating drill wells.
Structural geology is closely allied to engineering mechanics, fluid dynamics and
material science. There is, however, an important difference between engineering
mechanics and structural geology (Fig. 1.1). Most of the deformation processes in the
crust are very slow. As a result geological structures develop over long periods of time,
usually in millions of years. Further, most of the rocks are very highly heterogeneous
materials. Consequently, structures we see in nature are end products of very slow
deformation processes in highly heterogeneous materials. We try to interpret the
deformation process and the initial condition from the end product. In engineering
problems, one generally studies the effects of various deformation processes on
undeformed and relatively homogeneous materials.
Another serious problem in structural geology, indeed in most subjects in earth
sciences, is that we have to deal with incomplete, sometimes conflicting, data set.
Therefore, structural inferences are interpretative and non-unique, and require
validation. But unfortunately, there is a general lack of enthusiasm for validation of
structural inferences!

1
Initial Final
Engineering

Structural geology

?
Figure 1.1. The difference between engineering mechanics and deformation in
the crust. In engineering mechanics, the effects of known stress systems on
initially homogeneous undeformed materials are usually studied. In structural
geology we see the end product of deformation in rocks, which are mostly
heterogeneous material. We try to infer the deformation process and initial
condition from the end product.

1.1 Aspects of structural study

Three different aspects of deformed rocks are analyzed in structural geology, viz.,
geometric, kinematic and dynamic.

• Geometric analysis is the qualitative description of size, shape and orientation of


a structure. Determination of orientation of fold axis from dip/strike data is an
example of geometric analysis. Interpretation of large-scale folding from
outcrop-scale structural data and map pattern is another example of geometric
analysis. Stereographic projection is a powerful tool for geometric analyses of
structures.
• Kinematics is a branch of mechanics that treats motion in an abstract framework,
without any reference to force or mass. In structural geology, kinematic analysis
is a mathematical description of movement of material points during
deformation in a rock. The stress or the rheological properties of rocks are not
taken into account during kinematic analyses.
• Dynamic analysis involves understanding applied forces that produce
deformation in the rock. The palaeostress analysis is an example of dynamic
analysis. Dynamics also include how rocks are strained in response to imposed
stress.

2
Also, three different types of studies are undertaken by structural geologists, viz.,
observational, experimental and theoretical.

• In observational studies deformation features in rocks are studied at different


scales varying from submicroscopic through outcrop in the field to global
scales. Typically such studies involve description of the geometry of different
structures and their order of formation. Also, variations in orientation of
different structures seen in outcrops are studied in order to interpret the
geometry of map-scale structures.
• The main aim of experimental structural studies is to reproduce in the laboratory
the structures seen in nature in order to gain insight into different factors that
control the geometry of different structures. One problem with experimental
studies is that the strain rates of laboratory experiments are much faster than
expected in natural deformations. Therefore, experiments are commonly carried
out on analog materials.
• In theoretical studies different types of structures are numerically modelled
through the application of various physical laws and using analytical or
numerical methods.

We look at deformed rocks at different scales. Following terms are used to denote
approximate scale of observation of deformed rocks.

• Global: Structural features observed in the scale of the world. Mid-oceanic


ridges, subduction zones and orogenic belts are observed in global scale.
• Regional: Generally denotes a scale of the order of a physiographic province or
basin, such as Dharwar craton, Himalayan fold-thrust-belt, Satpura basin etc.
• Map or Macroscopic scale: Structural features seen on a map, which correspond to
an area much bigger than an outcrop. The area covered by a map may vary
from several tens of square meters, to several tens or hundreds of square km.
• Mesoscopic: Structural features observed in an outcrop or handspecimen.

3
• Microscopic: Deformation features seen in a thin section under an optical
microscope.
• Submicroscopic: Deformation features that cannot be seen even at the maximum
magnification of an optical microscope but can be seen with TEM, SEM etc.

Depending on how pervasive at a particular scale of observation, a structural


feature may be penetrative or non-penetrative (Fig. 1.2).

• Penetrative: We consider a structural element to be penetrative, or


homogeneously distributed, if the spacing of the structure is small compared to
the volume of the rock under observation.
• Non-penetrative: If the spacing of a structure is large as compared to the volume
of the rock under observation, then the structure is non-penetrative.

(a) (b)

Figure 1.2. The faulting is not penetrative in larger scale (a) because it is not
uniformly developed but in the scale of outcrop (b) the same faulting is
penetrative.

1.2 Geometry vs. strain vs. stress

Rocks accumulate permanent strain in response to an imposed stress during


deformation. The geometric features are the manifestations of permanent strain. Thus
in nature the sequence is stress → strain → geometry. Although everything starts with
stress, we do not observe stress directly. This is because stress is an instantaneous
quantity, i.e., it exists only at the moment it is applied. What we study in structural
geology is strain but in most rocks strain cannot be measured because strain markers
are uncommon. In deformed rocks, what we observe most of the time is geometry. We
4
can carry out geometric analysis in most of the deformed rocks but can carry out strain
measurements if we are very lucky to have strain markers. It is only in very rare cases it
is possible to do stress analysis.

5
6
2. Planes and Lines
The geometries of geological structures are described in terms of planar and linear
structural elements. Planar structural elements include bedding planes, axial surfaces
of folds, joint surfaces and fault surfaces. Hinge lines of folds, intersection lineation
produced by intersection of planar elements, mineral lineation and slickenside
lineation on fault surfaces are examples of linear structural elements. The following
terms are used to describe the attitude of planes and lines:

• Attitude: It is a general term used to indicate orientation of a structural plane or


line. The orientation is related to geographic coordinates, i.e., north-south and
east-west, and the horizontal. The attitude is specified in terms of bearing and
inclination.
• Bearing: It is the angle between a line and a specified geographic coordinate
direction, measured in a horizontal surface. The geographic coordinate
direction is usually north direction, but can also be east, south or west. Note
that an inclined line has to be projected onto a horizontal surface before bearing
can be measured.
• Inclination: It is the angle between a plane or a line and an imaginary horizontal
line, measured downward in a vertical plane.
• Strike: The line of intersection between a plane of interest and an imaginary
horizontal plane is called strike line. Note that the strike line must be a
horizontal line. The bearing of the strike line is the strike angle, or simply strike,
of the plane of interest.

7
N (North)
Trend Strike Strike line
D C
Plunge Pitch
Dip
A
B

Lin
e
H Plane
G

E
Strike line F

Figure 2.1. Block diagram showing different terms used to describe attitude of
planes and lines.

• Dip: Dip is the inclination of a plane of interest measured on a vertical surface


oriented perpendicular to the strike line. Dip (or true dip) is the maximum
inclination of a plane.
• Apparent dip: Inclination of a plane measured in any direction other than the
perpendicular direction to the strike line. The apparent dip is always less than
the dip (or true dip). Apparent dip in the direction of strike is always zero.
• Plunge: It is the inclination of a line from an imaginary horizontal line measured
on a vertical plane.
• Trend: It is the bearing of a line. If the line is inclined, then it is necessary to
vertically project the line onto an imaginary horizontal plane before trend can
be measured.
• Pitch: We measure pitch (or rake) of a line that lies on a plane. Pitch is the angle
between the line and an imaginary horizontal line, both lying on the same
plane.

Let us consider a block ABCDEFGH in which ABCD and EFGH are horizontal
planes and the other four planes are vertical (Fig. 2.1). In this block, CDEF is the plane
of interest containing a line DF. The direction from A to D is the geographic north
direction. The lines CD and EF are the strike lines for the plane CDEF. The line AB is
the vertical projection of EF onto the horizontal plane ABCD. Note that the strike lines

8
CD and AB are parallel though the strike lines are at different elevations. The distance
between AB and CD is a function of dip of the plane, higher the dip smaller the
distance. The ∠NDC is strike of the plane CDEF. Note that a strike line has two
bearings, 180° away from each other. The ∠BCF (or ∠ADE) is the dip (or true dip) of
the plane CDEF. The dip direction is D to A or C to B. The ∠BDF is one possible
apparent dip of the plane CDEF. The ∠BDF is the plunge of the line DF. Note that the
lines DB and DF are contained in the same vertical plane and DB is the vertical
projection of the line DF onto a horizontal plane. DB with bearing ∠NDB is the trend of
the line DF. ∠CDF is the pitch of the line DF measured from C side of the strike line
DC. It is important to remember that the terms strike and dip/dip direction describe
the attitude of planes, and trend and plunge give attitude of lines.

A
0
120
W E

Figure 2.2. A line AB makes 120°, measured clockwise, with the geographic
north. The bearing of the line can be stated equivalently as N120°, N300°, E30°S,
S60°E, N60°W or W30°N. Note that the suffix indicates the direction from
which bearing is measured.

The bearing of a line is usually stated in two different conventions. For example the
bearing of the line AB in Fig. 2.2 can be stated as N120° (i.e., 120° clockwise from N) or
E30°S (i.e., 30° from E towards S). It may also be stated equivalently as N300°, S60°E,
N60°W or W30°N.
Orientations of planes and lines observed in an outcrop can be measured directly
using a clinometer, which is a magnetic compass with provision to measure inclination.
The use of a clinometer is best demonstrated in an outcrop. The bearing of a line is

9
measured with respect to geographic North Pole (or true north direction) but the
magnetic needle in a clinometer obviously points towards magnetic North Pole. We all
know that true and magnetic North Poles are not same. In 2003, the magnetic North
Pole was located at about 78°18’N latitude and 104°W longitude near Ellef Ringes
Island, northern Canada and about 700 km from the geographic North Pole. The angle
between magnetic N-S line and geographic N-S line is called declination, which varies
from place to place on the surface of the earth. The magnetic north direction must be
corrected to get the direction of the geographic north. Most clinometers have provision
to make such a correction. The declination in India is close to zero and, therefore, we
can take magnetic north as determined with a clinometer as true north direction.
Two aspects of attitude of planes need to be remembered. Firstly, many a cross
sections are drawn on seismic reflection profiles. In some cases we have no choice with
our first cross section, such as in offshore or areas covered with alluvium or in desert.
In seismic reflection profiles, the horizontal axes are distance and the vertical scale is
two-way-travel time (TWT). Obviously, the horizontal and vertical scales are different,
and therefore, the “dip” of reflectors we see in the profiles are not true dips. Velocity
models are required in order to convert TWT into depth. Further, artifacts, such as
diffractions or velocity pull-ups/pull-downs may also give false dips of reflectors.
Structural interpretations based exclusively on seismic reflection profiles without depth
conversion and consideration for artifacts may give distorted picture. Secondly, we
prefer to draw structural cross sections perpendicular to the dominant orientations of
strikes of bedding planes/axial planes and/or trends of fold axes. However, more
commonly we find that in some parts, the dip directions are oblique to the line of
sections. In such cases, apparent dips should be used while constructing cross sections.
In order to understand the spatial relations between angular components shown in
Fig. 2.1 and defined above, three-dimensional visualizations of problems involving
orientations of planes and lines are very important. Once the ability to visualize in 3D
is developed, more efficient and quicker methods, including readily available
softwares, may be applied to solve real life structural problems.

10
D'
A D
K'
K
B N C d

d d P S
J
Q R
Plane P
I E H
M
F L G

Figure 2.3. Visualization of a three-point problem in a block diagram. The


problem here is to determine dip and strike of the shaded plane in the block.
See text for details.

Consider the shaded plane P, i.e., the plane AIGS, in Fig. 2.3. The plane occurs at
different “elevations” at the four edges of the block, i.e., at A, I, G and S at edges AE,
BF, CG and DH, respectively. We know that a strike line for a plane can be drawn if we
can find two points on the plane at the same elevation with respect to a datum plane
(e.g., mean sea level). On a geological map, we look for two intersections between a
rock contact and a topographic contour in order to draw a strike line. Points S and J are
such points and a line through them gives us a strike line at the elevation given by the
plane PQRS. The line KD is the projection of line JS onto the plane ABCD. Note that
along the strike line KD, the plane P is always at a depth of d. The line L′D′, which is
parallel to KD and passes through A, is also a strike line but at the elevation given by
the plane ABCD. Plane ANLE is a vertical plane oriented perpendicular to strike lines.
Line AM is the intersection between vertical plane ANLE and the plane P. Therefore,
∠NAM is the dip of the plane and AN (i.e., from A towards N) is the dip direction.
Block diagrams are useful to visualize problems involving angular components of
planes and lines, such as three-point problem, determination of strike/true dip from
apparent dips or determination of apparent dip in any direction from true dip/strike.
However, solving problems graphically with the help of 3-D diagrams is never easy.
The solutions may be obtained by projecting everything onto one plane. In Fig. 2.4, the

11
lines AI, AM and AS (traces of the plane along three vertical surfaces) have been
projected on the horizontal surface ABCD by rotating 90° along lines AB, AN, and AD,
respectively. This type of graphical construction may be employed to find solutions to
several types of problems involving dip and strikes of planes, for example: (1)
determination of true dip from apparent dips in the directions, such as, AB and AD, (2)
determination of apparent dips from true dip/dip direction, and (3) solution to three
point problems wherein dip/strike of a plane can be determined from known depths of
occurrence of the plane at different locations.

North
K'
A
J D'
I
d dip S

K d
B
N d
dip M D
direction

Figure 2.4. Construction of 2D projection diagram on a horizontal surface to the


3D problem shown in Fig. 2.3. See text for details.

The angular relationships between planes and lines can be readily determined
using a stereonet. These days almost all earth scientists use easily available computer
software for stereographic analyses of orientation data. However, it is extremely
important to be able to visualize in 3D the orientations of lines and planes in a
stereogram (Fig. 2.5). Consider that AB is the trace of an inclined plane P on a
horizontal outcrop surface and the line contains a line OL (Fig. 2.5a). If a sphere is
drawn centered at point O, then the plane P (and its extension above the surface)
intersects the sphere as a great circle (Figs. 2.5b). This great circle is a spherical
projection of the plane in 3D and must be projected onto 2D space before any
orientation problem can be solved. The equatorial circle in Fig. 2.5b is the surface on
which planes and lines are projected. In order to get the projection of a plane, lines are
drawn connecting each point on the part of the great circle in the lower hemisphere to

12
zenith point Z. Five such lines joining points 1 though 5 to Z are shown in Fig. 2.5c.
Each of these lines intersects the equatorial circle and the arc traced by the loci of such
intersection points is the projection of the plane P. The arc 1-2’-3’-4’-5 is the projection
of plane P and is called a great circle. A view down from Z will look like Fig. 2.5d and
is called a stereogram. The limiting circle is called primitive, which also represents a
horizontal plane. The line OL is projected as a point L′. Note that both the plane and the
line have lost one dimension after projection, as happens with any projection diagram.

Z Z

A N
W
S
O B E

N N
A 1
O W E
W E 2'
L' 3' 4'
B 2 s 5
L S
Plane P
L Plane P L 3 4
(a) (b) (c)

N Z N

1
Great
circle
2' N
W E Pole
W P' O E W E
L' O' dip dip
3' S
5
4'

(d) P (f)
S
(e)
S

Figure 2.5. Derivation of stereographic projection diagram. See text for details.

If a large number of planes are plotted as great circles, the stereogram may become
cluttered. In such a situation, planes are plotted as poles instead of as great circle. In
Fig. 2.5e line OP is perpendicular to plane P and is projected at point P' on the
equatorial plane. So, P' is pole to plane P. The primitive, i.e., the limiting circle
represents the great circle of a horizontal plane and a straight line passing through
center is a vertical plane. Therefore, great circles for planes with gentle dips will plot
closer to primitive and steeper planes will have their great circles closer to the center
(Fig. 2.5f). For poles it is just the reverse. Poles to planes with gentle dips will plot close
to the center, whereas the planes with steep dips will have their poles closer to

13
primitive. Note that if planes are gently dipping, their poles will have steeper plunge
and plunge of poles to steeply dipping planes will have gentler plunge. Similarly, lines
with gentle plunges will plot closer to primitive and steeply plunging lines will plot
closer the center. Using stereographic projections we can solve some of the three point
problems quickly and efficiently. This technique is also useful for statistical analyses of
orientation data.

14
3. Force and Stress
This section gives elementary descriptions of force and stress. For a more detailed

treatment, readers are referred to Ramsay (1967), Means (1976) or any recent textbook

on structural geology.

3.1 Force

Forces acting in rocks produce deformation structures that structural geologists


study. Therefore, it is necessary for us to understand how forces are distributed in the
earth and how they produce different types of structures. Since we usually study old
deformation, we do not see or measure forces directly in rocks because forces are
instantaneous. Rarely we can see or feel the effect of forces acting in the earth as at the
time of an earthquake. Although forces acting on rocks cause deformations in them,
structural geologists usually talk in terms of stress. Force and stress are not exactly the
same though they are closely related, as we will see.
High school physics textbook tell us that a force is an influence, which has an
intention to set a body at rest in motion or to change the velocity and direction of a
body in motion or to change the shape of a body. Note that a force may not be able to
do any of these but it has to have the intention. This, in essence, is the Newton’s first
law of motion.
Forces may be balanced or unbalanced. The forces are balanced if the summation of
all forces acting on a body is zero otherwise the forces are unbalanced. When balanced
forces act on a body it does not change its position at rest (or of uniform motion) and it
appears as if no force is acting on it. Unbalanced forces can move a stationary body or
they can stop/slow down/accelerate a moving body. Force is a vector quantity having

15
magnitude as well as direction. A force is called a body force if it can work from a
distance and depend on the amount of material affected. The gravitational force is a
body force. The forces that act across a surface of contact between two adjacent parts
are called surface forces. Tectonic forces, which drive lithospheric plates, are surface
forces. These two types of forces are closely related in the earth.
If a force F acting on a body of mass m produces acceleration a in the body then
according to Newton’s second law of motion (* indicates multiplication):

F=m*a

The unit of force in S.I. system is newton (N), which is the force acting on a body of
1 kg mass produces an acceleration of 1 meter (m) per second (s) per second. Therefore,

1 N = 1 kg * 1 m s-2

If we consider a force that produces an acceleration of 1 cm s-2 on a body of mass 1


g, then the unit is dyne (dyn):

1 dyn = 1 g * 1 cm s-2, and since 1 kg = 103 gm and 1 m s-2 = 102 cm s-2


1 N = 105 dyn

Most of the time we use the terms weight and mass interchangeably but they are
not the same. When a body of mass m is allowed to fall freely, its acceleration is that of
gravity g and the force acting on it is its weight W, so

W=m*g

The value of g varies from place to place, but for our purpose it can be taken to be a
constant with a value of 9.8 m s-2. So, a free falling body of mass 1 kg will have a weight
of

W = 1 kg * 9.8 m s-2 = 9.8 N

16
Therefore, the weight of a body is the gravitational force exerted on it by the earth.
Weight, being a force is a vector quantity. The direction of this vector is the direction of
the gravitational force, i.e., towards the center of the earth. So when we say weight is 1
kg, what we actually mean is the gravitation force exerted on a body of mass 1 kg!
Newton and dyne are wholly formal units of force but they are not familiar and
difficult to relate to everyday experiences. One kg, on the other hand, is very familiar
and is used frequently as somewhat informal unit of force. If we put a block of iron of
mass 10 kg on a tabletop, the force exerted by the iron block on the table is 98 N or 98 *
105 dyn. Alternately, we may say that the iron block exerts a force of 10 kg on the table.
Let us consider the uppermost cubic meter of a granite cube in an outcrop with the
cube separated from the surrounding rock by open joints on its four vertical sides. We
now wish to calculate the force acting on the basal surface of the granite cube. This
force will be sum of the force exerted by the atmosphere acting on the top surface of the
cube plus the force exerted by the cube. The mass of a column of atmosphere occurring
on top of the cube is 9700 kg and the mass of a cubic meter of granite (ρ = 2.7 g cm-3) is
2700 kg. The force acting on the base of the cube can be stated as 12400 kg or more
formally as:

F = (9700 + 2700) kg * 9.8 m s-2 = 121520 N = 121520 * 105 dyn

Let us consider a plane half way down the above-mentioned cube of granite. The
top half of the cube plus the atmosphere on the top will exert a force across the plane
and this force (10.83 * 104 N) can be represented by a vertically downward pointing
vector. According to Newton’s third law of motion, the lower half of the cube will exert
an equal and oppositely directed force. In this case the two force vectors will point
towards each other and particles on either side of the plane will be pushed closer
together. Such forces are called compressive force. If the force vectors point away from
each other, the particles on either side of the plane will be pulled away from each other
and the force is tensile force. Compressive forces are given positive sign and tensile
forces are given negative sign (Fig. 3.1b).

17
A force vector directed across a plane can have any orientation with respect to the
plane. If the force vector is oriented perpendicular to the plane it is a normal force (Fn),
if it is oriented parallel to the plane it is a shear force (Fs). A force vector may not be
oriented either normal or parallel to the plane. Such a force vector can be resolved into
normal force and shear force, just like any other vector (Fig. 3.1a). The shear force can
be further resolved into forces parallel to any convenient coordinate directions (Fig.
3.1a). Sign conventions for different kinds of forces are shown in Figs. 3.1b,c.

3.2 Stress

If we place a 10 kg weight on top of a cubic meter of quartzite nothing happens. But


if we put the same 10 kg weight on a grain of sand it may get pulverized. Intuitively we
can say that the grain of sand “felt” a lot more force than the cube of quartzite although
both of them were under the same force. In order to express this we need to define a
new term called stress.

z
A FBA A FBA
Plane P
F y Plane P (b) B FAB
Fn B FAB
Fx compressive (+ve) tensile (-ve)

Fs Fy
x Plane P
(a) (c) +ve -ve

Figure 3.1. (a) Resolution of force vectors. The total vector F (bold indicates
vector) acting on a plane P can be resolved into normal force (Fn) and shear
force (Fs). The shear force can further be resolved into force vectors parallel to x
(Fx) and y (Fy) co-ordinate axes. (b) Sign and notation conventions for surface
forces: compressive and tensile forces are considered +ve and –ve, respectively.
FAB indicates force exerted by body B towards body A and FBA indicates force
exerted by body A towards body B (c) Sign conventions for shear forces:
counter-clockwise and clockwise shear forces are +ve and –ve, respectively.

Let us again consider the uppermost cubic meter of granite cube in an outcrop with
the cube separated from the surrounding rock by open joints on its four vertical sides.

18
The force on the 1 m2 basal plane of the cube is 121520 N or 12400 kg. If we normalize
force by the area we get 121520 N m-2 or 12400 kg m-2. Note that these values are same
as 12.152 N cm-2 and 1.24 kg cm-2, respectively. In other words, we get the same
normalized force no matter what the area of the basal plane we take so long we have 1
m high rock column and the column of atmosphere on top. This normalized force, i.e.,
force divided by area, is called stress (σ):

Stress (σ) = force (F) / area (A)

Let us consider a situation where non-uniform forces are acting across a plane (P).
The stress across a small part of the plane (∆A) is given by

σ = ∆F / ∆A

If we take an infinitesimally small area we may consider it to be a point, p. The


stress across the plane P at point p will be given by

σ = dF / dA

Stress on a plane is a vector quantity because it is the product of a vector (∆F) and a
scalar (1/ ∆A). Stress has magnitude equal to the ratio of force to area and a direction
parallel to the force across the plane. The formal unit of stress is pascal (Pa):

stress = force / area = (kg m s-2) m-2 = N m-2 =Pa

In the earth, most stresses are significantly larger than a pascal, so we frequently
use megapascal (Mpa)

1 Mpa = 106 Pa = 10 bar = 9.8692 atm

Like any other vector, a stress vector can be resolved into components to any
convenient reference directions (Fig. 3.2). Obviously, stress vectors can be added
vectorially so long as the stress vectors are related to a single plane. Stresses acting

19
perpendicular and parallel to the plane are called normal stress and shear stress. A
stress vector acting across a plane P at point p has a stress vector of equal magnitude
but oriented in opposite direction. If the two vectors point towards each other the stress
is compressive (+ve), otherwise the stress is tensile (-ve). Note the similarities between
Figs. 3.1 and 3.2.

3.3 Stress ellipse

There is another meaning of the word stress. As discussed above, there are two
parallel but oppositely oriented stress vectors with the same magnitude at a point p on
a plane P. There may be infinite number of differently oriented planes, all of them
passing through point p. At point p and for each of the planes there are two parallel but
oppositely oriented stress vectors. A collection of all the stress vectors associated with
planes of all possible orientations but passing through the same point p is also called
stress - stress at a point. But this stress is not a vector but a tensor. The loci of heads (for
tensile, Fig. 3.3a) or either tails (for compressive, Fig. 3.4b) of all the stress vectors
acting at a point trace an ellipse in two dimension, called stress ellipse. It is obvious
that there is no stress ellipse for a situation where some of the stress vectors are
compressive, and others are tensile. One of the planes P with the associated stress
vectors are also shown in Figs. 3.3a and 3.4a, note that the stress vectors are not
perpendicular to the plane P. This is generally true and a stress ellipse should not be
thought of as collection of normal stresses. However, there are four stress vectors in an
ellipse those act across perpendicularly oriented planes. These vectors are coincident
with the major and minor axes of the stress ellipse. In other words, these are normal
stresses and are called principal stress axes. The planes across which principal stress
axes operate are called principal planes of stress (Figs. 3.3b and 3.4b).
Equation of a stress ellipse has the same form as any other ellipse. If in a stress
ellipse:
• the centre of the ellipse is centered at the origin, i.e., at x = 0 and y = 0 in x-y
coordinate system
• major and minor axes are coincident with x and y axes, respectively

20
• the major and minor radii of the ellipse are the magnitudes of σ1 and σ2,
respectively
then, the equation of the stress ellipse is (σx2 / σ12) + (σy2 / σ22) = 1

z A σBA A σBA
Plane P
σn σ y Plane P (b) B σAB σAB
B
σx compressive (+ve) tensile (-ve)

σs σy
x Plane P
(a) (c) +ve -ve

Figure 3.2. Resolution, and notation and sign conventions of stress vectors.
Note the similarity with Fig. 3.1. (a) The total stress vector σ (bold indicates
vector) acting on a plane P can be resolved into normal force (σ n) and shear
force (σ s). The shear force can further be resolved into force vectors parallel to x
(σ x) and y (σ y) co-ordinate axes. (b) Sign and notation conventions for surface
forces: compressive and tensile forces are considered +ve and –ve, respectively.
σAB indicates force exerted by body B towards body A and σBA indicates force
exerted by body A towards body B (c) Sign conventions for shear forces:
counter-clockwise and clockwise shear forces are +ve and –ve, respectively.

Principal
planes of
stress

B
σAB σ2
A
σ1 σ1
p Plan e P

σBA σ2
(a) (b)

Figure 3.3. (a) Stress ellipse given by loci of heads of all the stress vectors acting
at point p on plane P. σAB and σBA associated with plane P are also shown. Stress
vectors in this case point away from point p and thus they are tensile. Note that
σAB and σBA are not perpendicular to plane P. (b) Two pairs of stress vectors
oriented perpendicular to the planes across which they operate. These stress
vectors are called principal stresses and coincide with the major and minor axes
of stress ellipse. The planes oriented perpendicular to the principal stress
directions are called principal planes of stress.

21
Principal
planes of
stress

B
σAB σ2
A
σ1 σ1
Plane P
p
σBA σ2
(a) (b)

Figure 3.4. Stress ellipse for a situation where all the stress vectors are
compressive. See caption to Fig. 3.3.

3.4 Stress ellipsoid

A stress ellipsoid represents collection of stress vectors in three dimensions (Fig.


3.5), all of them acting at a point. It is exactly analogous to stress ellipse in two
dimensions except that there are three principal stress directions and three principal
planes of stress.

σ1
σ2
σ3

Figure 3.5. Stress ellipsoid, i.e., stress in 3D. Note that there are three principal
stresses and three principal planes of stress.

The three principal stresses are σ1, σ2, and σ3 with magnitudes σ1 > σ2 > σ3. Three
principal planes of stress define three stress ellipses containing σ1-σ2, σ2-σ3 and σ1-σ3
axes of the stress ellipsoid. One, two or three principal stresses may have non-zero
values. In triaxial state of stress all the three principal stresses have non-zero values, in
biaxial state of stress only two principal stresses have non-zero values and in uniaxial
22
state of stress only one principal stress has non-zero value. If σ1 ≠ σ2 ≠ σ3 it is called
general or polyaxial stress. In axial stress two of the three stresses have same magnitude.
If σ1 > σ2 = σ3 it is called axial compression and the ellipsoid is a prolate spheroid. In axial
extension σ1 = σ2 > σ3 and the stress ellipsoid is an oblate spheroid. In hydrostatic stress
magnitudes of all the three stresses are same, i.e., σ1 = σ2 = σ3. The ellipsoid is a sphere
and the stress is called pressure, which is the only kind of stress that can exist in a fluid
at rest. The equation of a stress ellipsoid is
(σx2 / σ12) + (σy2 / σ22) + (σz2 / σ32) = 1

23
24
4. Mohr Circle
Normal (σn) and shear (σs) components of stress on a real or imaginary plane are
important for the understanding of theories of development of faults and joints. The
equations of normal and shear stresses in terms of principal stresses (σ1-σ2, σ1-σ3 or σ2-
σ3 space) are most useful in understanding the basic concept of stress. Finally, pictorial
representation of how normal and shear stresses vary with the change in orientation of
the plane with respect to principal stresses is very illustrative; this we do with the help
of a diagram called Mohr circle.
The equations for the normal and shear stresses can be derived from force-balance
problems, which assumes that if a body is in equilibrium (i.e., it does not move or spin)
then all the forces in any one direction sum to zero. Note that we should always
balance forces and not stresses. We should first find the forces acting in any particular
direction, determine stresses in terms of the forces, and then derive the expressions for
normal and shear stresses.
Consider a prismatic body in the earth with two of the bounding planes oriented
parallel to the maximum and minimum force vectors (Fig.4.1). Another bounding plane
P has an area A and whose orientation may be specified by the angle θ made by the
perpendicular to the plane with maximum (F1) force direction. Therefore, the areas of
the left and right faces of the prism are Asinθ and Acosθ, respectively. Resolution of the
magnitudes (F1 and F2) of forces acting on the three bounding planes is shown in Fig.
4.1. Resolved forces are parallel and perpendicular to the plane P.
Forces trying to push the prismatic body upward must be equal to the forces trying
to push the prismatic body downward, if the body has to remain in equilibrium.
Therefore, from the balance of forces acting across the plane P (Fig. 4.1):

F1n + F3n = F1 cosθ + F3 sinθ

25
or, Fn = F1 cosθ + F3 sinθ (4.1)

And, from balance of forces acting parallel to the plane P we obtain:

F1s + F3 cosθ = F3s + F1 sinθ


or, F1s – F3s = F1 sinθ - F3 cosθ
or, Fs = F1 sinθ - F3 cosθ (4.2)

Fn
Plane P of area A
F1 F1n F1
F3 F3n
F3
F1s Fs F3s
θ
Perpendicular
to plane P

F3 cosθ θ
A

os F1 sinθ
s in

θ Ac
θ

F3 sinθ
F3 θ
F1 F1 cosθ
θ

Figure 4.1. A prismatic body bound by planes parallel to F1 and F3 directions,


and a plane P with area A and whose normal is inclined at θ to the F1 direction.
As the body is at rest, forces acting in any one direction must sum to zero. Force
trying to move the prism from left to right is exactly balanced by the force
acting from right to left and force trying to move the body from top to bottom is
balanced by the force acting from bottom to top. Resolution of magnitudes of F1
and F3 forces on the three bounding surfaces are shown on the figure.

We now calculate the stresses in terms forces from the relation, stress = force / area:

σn = Fn / A σs = Fs / A
or, Fn = A σn Fs = A σs (4.3)
and
σ1 = F1 / (A cosθ) σ3 = F3 / (A sinθ)
or F1 = σ1 A cosθ F3 = σ3 A sinθ (4.4)

26
Replacing (eq. 4.3) and (eq. 4.4) in (eq. 4.1) and (eq. 4.2), we obtain

σn = σ1 cos2θ + σ3 sin2θ (4.5)


σs = σ1 sinθ cosθ + σ3 sinθ cosθ (4.6)

Using the identities cos2θ = (1 + cos2θ)/2, sin2θ = (1 - cos2θ)/2 and sinθ cosθ = sin2θ in
eqs. 4.5 and 4.6, and collecting terms we obtain

σn = ½ (σ1 + σ3) + ½ (σ1 - σ3) cos2θ (4.7)


σs = ½ (σ1 - σ3) sin 2θ (4.8)

From the above two equations we can calculate σn and σs on any plane, given θ, σ1,
and σ3. The above equations are derived for the σ1-σ3 principal plane. We can derive
similar equations for σ1-σ2 and σ2-σ3 principal planes also.
Equations (4.7) and (4.8) define a circle with σn and σs as x and y axes, respectively.
The center of the circle (on the x-axis) and radius are given by (σ1 + σ3)/2 and (σ1 -
σ3)/2, respectively. This circle is known as the Mohr circle for stress.

σ1
325 bars

70O
σs, bars

70O 100
P Q

P
140O 40 O σn, bars
75 bars 20O
σ3 Q − 100 σ3 100 200 300 σ1 400

− 100

(a) (b)

Figure 4.2. Construction of a Mohr circle. (a) A very small rectangular block
showing state of stress and orientations of two planes on which normal and
shear stresses are to be determined using Mohr circle. (b) Mohr circle for the
state of stress as shown in (a). See text for details.

27
Let us consider a very small rectangular block in which there is a plane P
inclined at 70° from the horizontal. σ1 is vertical and has a magnitude of 325 bars,
and σ3 is horizontal with a magnitude of 75 bars. A line perpendicular to plane P
makes 70° with σ1 direction, i.e., θ is 70° (Fig. 4.2a). We want to show pictorially the
magnitudes of σn and σs on this plane; we can, of course, calculate the same using
equations 4.7 and 4.8. A pair of orthogonal coordinate axes is drawn with σn and σs
as horizontal and vertical axes, respectively (Fig. 4.2b). The two axes are graduated
in units of bars with same scale on both the axes. On the positive side of σn, points
are located at 325 and 75 bars and marked as σ1 and σ3, respectively. We locate a
point half way between σ1 and σ3, i.e., at 200 bars, i.e., (σ1 + σ3)/2. Taking this point
as center we draw a circle that passes through both σ1 and σ3. The resultant circle is
Mohr circle of stress for a state of stress with magnitudes of σ1 and σ3 at 325 and 75
bars, respectively. We find a point P on the Mohr circle such that a line drawn from
P to the center makes an angle of 140° (i.e., 2θ) from the positive end of the σn axis.
The coordinates of point P are 80.3 and 104.2 bars, which are the σs and σn,
respectively. The plane Q (Fig. 4.2a) inclined at 20° (θ = 70°) from horizontal has σs
and σn values of 80.3 and 295.8 bars, respectively. So it seems that every point on
Mohr circle represents a plane and the coordinates of the point equal normal and
shear stresses associated with the plane. Therefore, the circle is the loci of infinite
number of points, each of which represents stress on a plane whose orientation is
specified by θ.
Fig. 4.3 shows how the different terms in equations 4.7 and 4.8 are related to
coordinates of point P on the Mohr circle. Figs. 4.2, 4.3 are drawn for the principal
plane of stress containing σ1-σ3; this is two dimensional. We can also draw Mohr
circles for σ1-σ2 and σ2-σ3 principal planes. The three Mohr circles in Fig. 4.4
represent three-dimensional state of stress. The largest, intermediate and smallest
circles represent Mohr circles for σ1-σ3, σ1-σ2 and σ2-σ3 surfaces, respectively.

28
σn = [(σ1 + σ3)/2 + (σ1 − σ3)/2] cos2θ

+σs
σn P
1/2(σ1−σ3 ) sin2 θ
1/2(σ1−σ3 )
σs
−σn σ3 σ1 +σn

1/2(σ1+σ3 )
1/2(σ1−σ3 ) cos2θ

−σs

Figure 4.3. Mohr circle showing significance different terms in equations 4.7
and 4.8.

σs
σ1-σ3
σ1-σ2
σ2-σ3
−σn σ3 σ2 σ1 σn

−σs

Figure 4.4. Mohr circles for state of stress in three dimensions. The three circles
represent three principal planes of stress.

Sign conventions for the Mohr diagram are as follows: compressive normal stresses
and shear stresses represented by counterclockwise or sinistral pair of arrows are
positive and tensile normal stresses and shear stresses represented by clockwise or
dextral pair of arrows are negative (Figs. 4.2, 4.3); θ is positive for planes whose normal
can be located towards counterclockwise direction from σ1, otherwise θ is negative (Fig.
4.5).

29
Plane P Plane Q
zP

σ1 σ1
−θ
zP
(a) (b)

Figure 4.5. The sign convention for angle θ. (a) +ve θ. (b) –ve θ.

Two perpendicular planes (P and Q, Fig. 4.6) with θ = ± 45° (or 2θ = ± 90°) have
the largest absolute magnitude of shearing stress, which is equal to the radius of the
Mohr circle [or, (σ1 - σ3)/2 for σ1-σ3 principal plane]. Shear stresses for planes
oriented perpendicular to any of the principal stress directions are zero. It may
sound logical to presume that shear fractures (i.e., faults) should form at ± 45° to σ1.
As we will see (section 7) shear fractures usually develop at angles less than this.

σs

σ3 P

P Q
(σ1−σ3)/2
−σn O σn
σ1 σ3
+ 90
σ1
- 90O

−(σ1−σ3)/2

(a)
(b) Q
−σs

Figure 4.6. Maximum shearing stress is possible on planes oriented at ± 45° (2θ
= ± 90°) to σ1 axis.

Different classes of two-dimensional state of stress at a point are shown in Fig.


4.7. In hydrostatic tension (lithostatic tension for rocks), stress across all planes is
tensile and equal and the Mohr circle is a point on the negative side of the σn axis
(Fig. 4.7a). If the stress across all planes is compressive and equal, the state of stress
may be called hydrostatic compression and the Mohr circle is a point on the positive
side of the σn axis (Fig. 4.7b). The term “hydrostatic” refers to the stress experienced
by a fluid at rest. However, this term is also widely used to describe similar state of

30
stress in solid. Both the principal stresses can be either positive or negative and the
states of stress are general tension (Fig. 4.7c) or general compression (Fig. 4.7d),
respectively. In uniaxial tension only one principal stress is non-zero and it is tensile
(Fig. 4.7e), whereas in uniaxial compression only one principal stress is non-zero and
it is compressive (Fig. 4.7f). In many states of stress one principal stress is tensile
and the other principal stress is compressive (Fig. 4.7g,h,i). Pure shear stress is a
special class of state of stress where the two principal stresses have the same
magnitude but opposite sign (Fig. 4.7i). In such cases, planes of maximum shear
stress are also planes of pure shear, i.e., normal stresses across these planes are zero.
Except for lithostatic tension (Fig. 4.7a), all other classes of stress are possible in the
earth.

σs

c e g i h f d
−σn a b σn

−σs

Figure 4.7. Mohr circles for various types of state of stress in two dimensions.

Any non-hydrostatic state of stress, either in two-dimensions or in three-


dimensions, can be decomposed into two parts: a mean stress (σm) and a stress deviator
or deviatoric stress (σd). The mean stress is the average of the principal stresses:

σm = (σ1 + σ2)/2 in two dimensions, and


σm = (σ1 + σ2 + σ3)/3 in three dimensions.

The deviatoric stress is defined as:


σd = σn - σm

31
and is a measure of how much the normal stress in any direction deviates from the
mean or hydrostatic stress. Along three principal stress directions we have three
principal deviatoric stresses whose magnitudes are given by:

σ1d = σ1 - σm
σ2d = σ2 - σm
σ3d = σ3 - σm

Note that sum of the right hand terms in the above equations is zero. For
example, a state of stress with the values of three principal stresses 750, 1050, and
1560 bars can be thought of as combining a mean stress of 1120 bars and three
deviatoric stresses of –370, -70, and 440 bars. Mean stress in two-dimension is a
state of hydrostatic tension or compression and locates the center of the Mohr circle;
deviatoric stress is a pure shear stress. Given mean stress and deviatoric stress we
can construct the Mohr circle.
The deviatoric stress is responsible for distortion (or strain) in a body. The
distortion may be elastic (i.e., reversible) or plastic (i.e., permanent). Most
deformations are result of differential stress rather than the absolute magnitudes of
principal stresses, except for dilation. Differential stress is the difference between
the magnitudes of maximum and minimum principal stresses (σ1 - σ3). The mean
stress can be thought of as hydrostatic (or isotropic) part of the stress system and
causes only volumetric changes (dilation) in the material. Mean stress also controls
the strength of materials. For example, fracturing is inhibited with increasing mean
stress.

32
5. Strain
5.1. Deformation

When force is applied to a rock body, the particles within the rock body are
displaced and the rock is said to be deformed. Two types of deformations are usually
recognized (Fig. 5.1).

• Homogeneous deformation: If particles arranged in straight lines in an


undeformed rock remain so after deformation, then the deformation is called
homogeneous. Another definition of homogeneous deformation is that all
parallel lines of particles remain parallel after deformation.
• Inhomogeneous deformation: In this type of deformation, straight lines of particles
become curved after deformation and parallel lines of particles loose their
parallelism after deformation.

(a) (b) (c)

Figure 5.1. Homogeneous and inhomogeneous deformations. (a) Original


undeformed grid. (b) Homogeneous deformation wherein straight lines remain
straight and parallel lines remain parallel after deformation. (c) If straight lines
become curved and parallel lines loose their parallelism, the deformation is
inhomogeneous.

Mother Nature does not draw a grid, as in Fig. 5.1a, before deforming a rock for our
convenience! In some cases, however, Nature preserves features that can be used to
determine type and amount of deformation. For example, the branches in plant fossil

33
Neuroteris are approximately straight and parallel before deformation (Fig. 5.2a). The
straightness and parallelism may be preserved (Fig. 5.2b) or destroyed (Fig. 5.2b) after
deformation indicating homogeneous or inhomogeneous nature of deformation,
respectively.

(b)
(a) (c)

Figure 5.2. Undeformed (a) plant fossil Neuroteris showing homogeneous (b)
and inhomogeneous (c) deformations

Four independent geometric processes contribute to the total displacement of


particles during deformation:

• Rigid-body translation is the movement of the entire body through space in such
a way that the shape does not change. The movement vectors for all the
particles in any external coordinate system have the same orientation and
magnitude (Fig. 5.3a).
• Rigid-body rotation also involves movement without any change in shape.
However, in this case the body rotates about a single point, which is fixed with
respect to an external reference frame (Fig. 5.3b).
• Distortion produces change in the shape of the body due to movement of
particles with respect to each other (Fig. 5.3c).
• Volume change, as the term implies, change in volume of the body without any
change in shape. Volume change is also called dilation, although volume change
can be either positive or negative (Fig. 5.3d).

34
(a) (b) (c)

(d) (e)

Figure 5.3. Trilobite Phillipsia Geometric processes leading to different types of


deformations, as shown by trilobite Phillipsia. (a) Rigid-body translation. (b)
Rigid-body rotation. (c) Distortion, i.e., change in shape. (d) Dilation, i.e.,
volume change. (e) A combination of all the four types of deformation.

Note that the descriptions of distortion and volume change do not require an
external reference frame. The four processes are not mutually exclusive – Fig. 5.3e
shows a deformation which includes all the four processes. Distortion and dilation
together make up strain, which involves movement of particles relative to each other. In
a more quantitative sense, strain is a mathematical description that relates the size and
shape of a body before and after deformation. A rock body may undergo rigid-body
movement, either translation or rotation or both, but it is almost impossible to
determine the exact amount of rigid-body movement. However, the strain in a rock can
be precisely determined if the rock contains objects of known, original shape and/or
size (e.g., Figs. 5.2, 5.3. Note that deformation and distortion (and strain) are not exactly
the same although it is not uncommon to find them used interchangeably in the
literature. Further it is important to remember that translation/rotation of a rock body
may or may not be accompanied by internal strain, i.e., distortion/dilation.
There are several ways strain can be measured but all of them involve
measurement of some kind of change from an initial undeformed to a final deformed
state. The changes that are generally measured are changes in lines, angles and volume.

35
5.2 Change in line length

Changes in line lengths, called longitudinal strain, can be measured in different


ways, viz., extension (e), stretch (S), quadratic elongation (λ), and logarithmic or
natural strain (ε). If Li is the initial undeformed length of a line, Ld is the final deformed
length of the same line, and ∆L is the change in the length of the line then (Fig. 5.4):

e = (Ld – Li) / Li = ∆L / Li
S = Ld / Li = (1 + e)
λ = (Ld / Li)2 = (1 + e)2
ε = loge (Ld / Li) = loge (1 + e)

Elongation can be either positive or negative depending on whether a line has


extended or shortened. Stretch is always positive whether a line has extended or
shortened. It has a value of 1.0 if there is no change in length of a line, S < 1.0 for
shortening and S > 1.0 for extension. All the four parameters for longitudinal strain are
dimensionless. They are not independent, if we know one we can calculate the others;
which one to use for a particular problem depends entirely on convenience. However,
logarithmic strain is realistic for several reasons. For example, if one line contracts to
half of its original length and another line expands twice its original length, the
elongations of the lines are 0.5 and 1.0, respectively. For the same deformed lines
logarithmic strains are -loge2 and + loge2. For very large shortening of a line, elongation
tends towards –1.0, but the logarithmic strain approaches -∝. The stress-strain curves
for isotropic materials are straight (i.e., linear) if logarithmic strain is used.
In nature we almost always measure Ld from deformed linear objects, such as
boudinaged quartz vein (Fig. 5.4b). We can put the boudins back into their original
position and determine Li, assuming no volume change. Note that in Fig. 5.4a the
length of the line has decreased (shortened) but the line in Fig. 5.4b has increased
(extended). In both the cases the change in the length of the line is 2.46 mm but
parameters describing longitudinal strain are different.

36
Undeformed (length Li) Deformed (length Ld )
∆L ∆L

Deformed (length Ld ) 2 cm
Undeformed (length Li)
(a) (b)

Figure 5.4. Longitudinal strain. (a) A line with initial length Li = 9.15 cm has
shortened to 6.69 (= Ld). The elongation (e), stretch (S), quadratic elongation (λ),
and natural strain (ε) are -0.27, 0.73, 0.53 and -0.31, respectively. (b) A more
practical scenario wherein boudinaged quartz vein can be used to determine
both Li and Ld. The e, S, λ and ε are 0.37, 1.37, 1.87 and 0.31, respectively.

5.3 Change in angle

Shear strain is a measure of change in angle between two originally perpendicular


lines. Consider a rectangle ABCD, which after deformation becomes a parallelogram
A’B’CD (Fig. 5.5). The angle between lines AD and CD has changed from 90° (∠ADC)
before deformation to α (∠A’DC) after deformation. We can state the shear strain in
two different ways (Fig. 5.5):

• Angular shear (ψ): This gives the change in angle, i.e., ψ = 90° - α.
• Shear strain (γ): This represents displacement (distance x) of a particle at a
distance y from a particle that does not move. From Fig. 5.5:
γ= tanψ = x/y
or, x = y tanψ = y γ
if y is unit distance then
x = γ = tanψ

Shear strain can be determined if appropriate markers are present in rocks. For
example, we can determine shear strain from deformed trilobite Phillipsia because of
inherent bilateral symmetry (Fig. 5.6). Similarly, well-preserved worm burrows or mud
cracks and stratification surfaces can be used to determine shear strain.
The original perpendicular line (e.g., line AD in Fig. 5.5) may move either in a
clockwise direction or in an anticlockwise direction with respect to the original

37
orientation. Clockwise and anticlockwise shear strains are given negative and positive
signs, respectively.

x
A A’ B B’

y ψ

D C

Figure 5.5. Shear strain illustrated by a rectangle ABCD that has changed into a
parallelogram A’B’CD after deformation. ∠ADA’ or ∠BCB’ is the angular shear
(ψ). Shear strain, γ = tanψ = x/y. If y is of unit length, γ = x.

(a) (b)

Figure 5.6. Trilobite Phillipsia shows shear strain. Morphology of Phillipsia is


such that two mutually perpendicular imaginary lines can be drawn (a). These
lines can be used to determine shear strain from a deformed fossil.

5.4 Change in volume

Volumetric strain or change in volume during deformation is called dilation (∆). If


Vi is the initial volume and Vd is the volume after deformation and ∆V is change in
volume after deformation, then dilation is given by:

∆ = (Vd - Vi) / Vi = ∆V / Vi

Although to dilate is to enlarge, dilation can have positive (i.e., enlarge) or negative
(i.e., contract) values. In two dimensions, we can only determine change in area.

38
5.5 Strain ellipse and ellipsoid

Strain ellipse (in 2D) and ellipsoid (in 3D) are elegant ways to depict homogeneous
deformation of a body as a whole. If particles lying on the periphery of a circle are
subjected to homogeneous deformation, the particles will trace an ellipse after
deformation. This ellipse is called a strain ellipse. In three dimensions, a sphere in the
undeformed state turns into an ellipsoid in the deformed state.
Let us look at the strain ellipse from a different viewpoint (Fig. 5.7). A circle
describes a collection of straight lines with equal length but of different orientations, all
passing through one point, which is the centre of the circle. Each of the lines connects
two particles on either side of the circle. During deformation particles will move with
respect to each other and the length of the lines will change. Stress ellipse (or ellipsoid
in 3D) describes a collection of straight lines in deformed state, all passing through the
same point, which is the centre of the ellipse. Obviously, there is no stress ellipse for
inhomogeneous deformation because straight lines do not remain straight.

z
z
λ1
1 λ3

1 x x

(b)
(a)

Figure 5.7. Strain ellipse. See text for discussion.

If the radius of the initial circle is taken to be of unit length, the major and minor
axes of the ellipse can be represented by √γ1 (= 1+e1) and √γ3 (= 1+e3), where √γ1 and √γ3
are maximum and minimum elongations (Fig. 5.7). So, the equation of strain ellipse
centered at origin is,

x2/γ1 + z2/γ3 = 1

39
Similarly, the equation of strain ellipsoid is

x2/γ1 + y2/γ2 + z2/γ3 = 1

where, γ1 > γ2 > γ3. The three elongations directions are usually taken parallel to x, y,
and z co-ordinate axes, respectively. Plane strain is a type of deformation where γ2 = 1,
i.e., along the intermediate axis of the strain ellipsoid there has not been any shortening
or elongation.

5.6 Finite and infinitesimal strain

When we look at a diastrophic structure in nature, such as a fold or a distorted


fossil, we know that the rock has undergone some amount of strain. However, it is
important to remember that the strain that we may observe and measure in rocks did
not develop instantaneously but accumulated in small increments over a period of
time. This is because, like most natural processes, deformations are also very slow.
Therefore, we observe the end product of a series of deformed states and straining of
rocks should be considered as progressive deformation. The final state of strain is
called finite strain and small incremental strains are known as infinitesimal strains. It is
possible that a line that shows finite extension may have undergone shortening at some
stage progressive deformation.

Shortening

Lines of no
finite elongation

Extension

Figure 5.8. Initial undeformed circle superimposed on strain ellipse. Two lines
can be drawn by joining opposite points of intersection between the circle and
ellipse. The lengths of these two lines have not changed during deformation.
They are termed as lines of no finite longitudinal strain and separate area where
all the lines have undergone finite extension (shaded area) from the area where
all the lines have undergone finite shortening.

40
5.7 Pure and simple shear

During a progressive deformation, all successive incremental strain ellipses may


have the same orientation, i.e., the axes of the strain ellipses remain parallel. Strain is
considered non-rotational and this type of strain is called pure shear (Fig. 5.9a). If the
strain ellipses change their orientations during progressive deformation, it is called
simple shear.

λ1
λ1
λ3
λ3

(a) (b)

Figure 5.9. (a) Non-rotational strain or pure shear. The major and minor axes of
the successive strain ellipses are coincidental. (b) Rotational or simple shear.
The strain ellipse undergoes rotation with respect to an external fixed reference
frame during progressive deformation.

41
42
6. Stress-Strain Relations
If a material is stressed, it gets strained. Rheology is the material response to applied
stress. Much of the classical theoretical and experimental principals of stress-strain
relations were developed in material science where materials are usually taken to be
homogeneous and isotropic. The behaviour of natural earth materials, i.e., rocks and
minerals, is often extremely complex because the rocks are neither homogeneous nor
isotropic. Even if a rock is approximately homogeneous to start with, it develops
fabric(s) during deformation and becomes heterogeneous. Nevertheless, stress-strain
relationships for homogeneous and isotropic materials can be used as a first
approximation for rocks. In the simplest form, materials respond to stress in two
different ways. When the stress is withdrawn, the material may return to original shape
and size, and strain is said to be recoverable. Otherwise the strain is permanent.
We can give stress parallel to the axis of a cylindrical rock and measure strain. The
results can be plotted on a stress-strain graph. At the initial stage stress-strain curves
are usually straight with steep slopes (Fig. 6.1a). The straight line implies that there is a
constant ratio between stress and strain and the steep slope means that small strain
accumulates for large incremental increase in stress. The most important aspect of this
part of the stress-strain curve is that the strain is completely recoverable, i.e., the
material returns to original dimension (zero strain) when stress components all drop to
zero. This kind of material behaviour is called elastic. The area under the straight line
curve is a measure of stored elastic energy. In some elastic materials, there is no time
lag between change in stress and corresponding change is stress. In other words, the
linear relation between stress and strain is instantaneous. This material behaviour is
known as Hookean behaviour. It is obvious that the Hookean behaviour and elasticity
are not exactly same. The following parameters are used to describe the properties of
elastic materials (see Fig. 6.1b):
43
• Young’s Modulus: In Fig. 6.1b the elongation parallel to the axis of the cylinder e
is given by (ld – li)/li. The stress is proportional to the strain for elastic material.
The proportionality constant, E, is called Young’s Modulus.

E=σ/e

This is for simple shortening and extension. The Young’s Modulus describes
how difficult it is to give longitudinal strain in a material. Higher the absolute
value of E, it will be more difficult to extend or shorten a material. Note that E
has same dimensions as stress (e.g., stress).

σ
Initial

Deformed

Stress, σ li ld
eL = (l d - l)/l
i i

Area is a eT = (wd - wi)/wi


measure of stored
elastic energy

wd
(a) Strain, ε (b) wi

Figure 6.1. Stress-strain diagram for elastic deformation. The area under the
curve represents stored elastic energy. (b) Longitudinal (eL) and transverse (eT)
elongations.

• Poisson’s Ratio: If eL and eT are longitudinal and transverse elongations,


respectively, then from Fig. 6.1b:

eL = (ld – li) / li and eT = (wd – wi) / wi

The Poisson’s Ratio (ν) is simply the ratio between eT and eL:

ν = eT / eL

44
So, ν is a dimensionless quantity. For small strains, the change in volume (∆) is
given by:

∆ = (σ / E) (1 - 2ν)

For constant volume deformation ν has a value of 0.5 no matter how high the
stresses get. If volume decreases with compressive strain or increases with
tensile strain, ν will have value less than 0.5. For most rocks ν varies between
0.25 and 0.33. Constant volume deformation indicates incompressible material.
For porous rocks with or without fluids volume change may be significant
during deformation.
• Rigidity modulus: If a cube of elastic material is subjected to shear stress σs it may
undergo shear strain γ, which is directly proportional to σs. The proportionality
constant is called rigidity modulus, modulus of rigidity or shear modulus (G).
Thus

G = σs / γ

Like Young’s Modulus, rigidity modulus also has the dimensions of stress.
Shear strain induces change in shape of a body. Therefore, rigidity modulus
describes resistance of an elastic body to change in shape.
• Bulk Modulus: This elastic property given the relationship between change in
pressure [∆P, i.e., hydrostatic stress, σmean = (σ1 + σ2 + σ3)/3] and consequent
volume change or dilation (∆) of a block of elastic material. Thus the Bulk
Modulas, K, is given by:

K = σmean / ∆

The inverse of Bulk Modulus is known as compressibility. Obviously, the Bulk


Modulus is a measure of ease or difficulty with which an elastic material can be
compressed.

45
The above elastic moduli are constant throughout isotropic materials but they vary
from place to place in anisotropic rocks geologists have to deal with. Further, the four
elastic moduli are not independent of each other but are related to each other through
some simple relations:

G = E/[2 (1 + ν)] = [3 K (1 + 2ν)]/[(2 (1 + ν)]

If stress is continued to be increased, the rock may suddenly fracture in the elastic
range (Fig. 6.2a) and the stored elastic energy is released in the forms such as elastic
waves, sound and heat. This is known as brittle failure and the deformation is called
brittle deformation. The value of the stress at which brittle failure takes place is called
brittle strength, rupture strength or fracture strength.
Under many conditions, including conditions of common laboratory experiments,
the rocks undergo permanent strain instead of fracturing if elastic limit is exceeded. If
stress is withdrawn strain is not completely recovered. The materials showing such this
behaviour of permanent straining are plastic materials and the deformation is called
plastic deformation or ductile deformation. The point on the stress-strain curve where
the changeover from elastic to plastic deformation takes place is called yield point,
which can be difficult to locate precisely during an experiment. The stress at the yield
point is called yield stress, yield-point stress or yield strength. Yield stress is not a
constant for any one rock type but varies with temperature, confining pressure, fluid
pressure and strain rate. Two kinds of deformations can be envisaged above elastic
limit:

• Perfect plastic: The material may continuously accumulate permanent strain at


the yield stress, i.e., the stress-strain curve will have zero slope (Fig. 6.2b). The
material is then called perfectly plastic.
• Strain hardening: In most deformed rocks, the stress-strain curve above yield
point has a positive slope (Fig. 6.2c). The slope of the curve is gentler than that
of elastic region and the curve is markedly non-linear. Strain can increase in this
situation only if stress is raised above initial yield stress; this process is called

46
strain hardening or work hardening (Fig. 6.2c). This strain-hardened deformation
is usually termed as plastic or ductile deformation. The rocks in this realm seem
to flow somewhat like a fluid and a plethora of structures develop. However, it
must be remembered that rocks are not fluid like water, which cannot sustain
shear stress at rest. If stress is withdrawn after some amount of plastic
deformation, such as at point X in Fig. 6.2c, the curve falls on the strain axis in
an approximately linear fashion (XY) giving the total amount of permanent
strain (OY, Fig. 6.2c) . If the stress is immediately reapplied, the curve goes back
along the linear path (YX) to the original position on the plastic deformation
curve. With increase in stress the curve continue along the original plastic
deformation path. The strain hardening may be thought of as elastic part of the
curve and yield strength moving continuously towards the right hand side of
the stress-strain diagram. With continued increase in stress the rock finally
ruptures at a point known as ultimate strength (Fig. 6.2c).

plastic
(strain hardening)
ultimate
perfect strength
brittle yield plastic yield X
strength strength strength rupture
Fracture

Stress, σ Stress, σ Stress, σ


elastic
elastic elastic

permanent strain

(a) (b) (c) Y


Strain, ε Strain, ε Strain, ε

Figure 6.2. Stress-strain relationships of elastic, perfect plastic and plastic (strain
hardened) behaviour of materials.

47
48
7. Brittle Fracture Criteria
Fracturing is an important process of rock deformation, particularly in the upper
part of the crust. Fractures are surfaces along which rocks loose cohesion, i.e., rocks
break along fracture surfaces. There are two types of fractures depending on the
relative motion between the two sides of a fracture. If there is no relative motion the
fractures are called joints.

• Extension fracture: In this type of fracture, relative motion between the two sides
is perpendicular to the fracture surface. They signify overall extension
perpendicular to the fracture surface. This type of fracture is also called Mode-I
fracture (Fig. 7.1a).
• Shear fractures: In this type of fracture, relative motion between the two sides is
parallel to the fracture surface. This type of fracture implies shearing movement
parallel to the fracture surface. The direction of relative motion may be
perpendicular or parallel to the edge of the fracture surface, they are called
Mode-II (Fig. 7.1b) and Mode-III (Fig. 7.1c) fracture, respectively. A mixed mode
fracture has components of both extension perpendicular to fracture and
shearing parallel to fracture.

(a) (b) (c)

Figure 7.1. Types of fractures depending on relative motion between two sides.
(a) Mode-I extension fracture. (b) Mode-II shear fracture. (c) Mode-III shear
fracture.

49
We also recognize two broad types of fractures in rocks in macrosopic scales
depending on quasi-mechanical behavior of rocks at the time of fracturing. They are
brittle and ductile fractures formed during brittle and ductile deformations,
respectively (Fig. 7.2).

• Brittle fracture: Brittle deformation occurs at the rupture strength in the elastic
regime with the formation of brittle fractures, which are surfaces across which
material loses cohesion. Rocks do not show any change in shape such that the
broken pieces fits together to give the original shape and size of the rock body
(Fig. 7.2b).

(a) (b) (c) (d)

Figure 7.2. Brittle and ductile deformations. (a) Undeformed cylinder. (b) Brittle
fracture. (d) Ductile deformation. (d) Brittle-ductile deformation.

• Ductile fracture: In ductile deformation, the rocks show permanent strain that
smoothly varies through the deforming material and no clear cut fracture
develops in macroscopic scale (Fig. 7.2c). This term does not signify any specific
deformation mechanism. Also, ductile deformation need not necessarily signify
plastic deformation mechanism. It is a general term for macroscopic flow of
rocks that can be accomplished by brittle deformation or plastic deformation or
a combination of both. If the ductile deformation is accomplished by fracturing
and rotation of individual grains or grain aggregates, the deformation is called
cataclastic flow. If the ductile deformation is dominated by flow of individual
grains through dislocation glide and climb, and diffusion, the deformation is
called plastic flow. After some amount of ductile deformation, a deforming rock
may fail and develop fractures. This type of fracture is called ductile fracture,

50
which may be considered as transitional behavior between brittle fracture and
plastic flow (Fig. 7.2d).

The type of fracture that may develop depends on the specific state of stress acting
on a rock body (Fig. 7.3):

• Hydrostatic compression: Rocks can withstand unlimited hydrostatic compressive


stress with volume change as the only manifestation of deformation (Fig. 7.3a).
This type of state of stress is possible in the earth, especially at great depths.
• Hydrostatic tension: Rocks break apart into random pieces if they are subjected to
high enough hydrostatic tensile stress when tensile stress exceeds cohesive
stresses that hold the particles in a rock together (Fig. 7.3b). This is a very
unlikely state of stress in the earth.
• Uniaxial compression: If σ2 and σ3 are zero or close to zero, as in uniaxial
compression, longitudinal splitting takes place parallel to compressive σ1 (Fig.
7.3c). The fractures formed by longitudinal splitting tend to be more irregular in
orientation and shape that other extensional fractures.
• Uniaxial tension: If σ1 and σ2 are zero or close to zero, as in uniaxial tension,
tension fractures form perpendicular to tensile σ3 (Fig. 7.3d).
• Axial tension: Extension fractures can also form perpendicular to σ3 if σ1 = σ2 >
σ3 > 0. Note that σ3 here is compressive (Fig. 7.3e).
• Axial or confined compression: Under this state of stress conjugate shear fractures
oriented at angles less than 45° to σ1 form (Fig. 7.3f). The two fractures do not
form simultaneously but form sequentially. In homogeneous material, it is not
possible to predict which one of the two will form first. The line of intersection
between two fractures can have any orientation on a plane perpendicular to σ1.
The possible orientations of shear fractures are tangent to a cone whose axis is
σ1. Displacements are parallel to fracture surfaces.
• Triaxial stress: Conjugate shear fractures form, as in axial compression. But in
this case, the intersection between the shear fractures is parallel to intermediate

51
stress axis σ2. The diehedral angle between the conjugate shear fractures is less
than 90°, which is bisected by σ1.

+σs +σ s

σ1 = σ 2 = σ 3 σ1 = σ 2 =σ 3
+σn +σ n
p (pressure)

(a) (b)
+σs σ1 +σ s
σ3

σ2 = σ3 = 0
σ1 +σ n σ3 σ1 = σ2 = 0 +σ n

(c) (d)

+σs σ3 +σ s σ1 σ1

σ1 σ 2 = σ3

σ3 +σ +σ n
σ1 = σ2 n σ 2 = σ3 σ1

(e) (f)

+σ s σ1

σ3

σ3 σ2 σ1 +σ n
σ2

(g)

Figure 7.3. Types of fractures depending state of stress in a rock body. In each
diagram state of stress is shown by a Mohr diagram and orientations of
fractures related to stress axes are shown. (a) Hydrostatic compression. (b)
Hydrostatic tension. (c) Uniaxial compression. (d) Uniaxial tension. (e) Axial
tension. (f) Axial or confined compression. (g) Triaxial stress.

Faulting is a consequence of shear stresses in the crust. Therefore, deviatoric stress


components in a stress system must have significantly large values for faults to
develop. It is important to understand the stress conditions for brittle failure of rock
because this gives us an insight on how faults and fractures develop during orogenic
52
processes. Laboratory experiments on rock deformation using triaxial testing apparatus
have been most illustrative in this regard. A schematic diagram showing the basic
features of such an experimental apparatus is shown in Fig. 7.4a. In more sophisticated
equipment pore fluid pressure and temperature can also be applied and controlled. The
three principal stresses can be controlled but two of them are equal. The forces applied
are perpendicular to the sample surfaces so that the principal stresses are either parallel
or perpendicular to the axis of the cylindrical sample.

Axial load
σa (σ1 or σ 3)
Steel
housing
σ14
Piston
Weak, impermeable
rubber/copper jacket
4 σs
σ13
(σa − σc), kbar

Fracture
3

Cylindical
rock σc
Confining
pressure
σ12 Fracture
sample (by a fluid) 2
σc σ11 σ12 σ13 σ14
σc = σ 2 = σ3
σn
σ11 1 2 3 4 5
kbars
σc = σ 1 = σ2 1

σa
(a) (b) 0 ε0 -5 -10 (c)
Axial strain, ε x 10
-3

Coulomb
fracture
σs behaviour
Mohr/fracture/ +σs φ
+σs Unstable failure envelope
p Fracture Transitional
0
θ = (90 +φ)/2
Stable
tensile
behaviour
p

+2θ +2θ
σc σ1 −σn +σn Τ0
σn −σn +σn
1 2 3 4 5 σ3 σ1
−2θ
kbars
Tensile −2θ
fracture
behaviour
p’
p’ −σs
(d) (e) (f) −σs

Figure 7.4. Derivation of fracture envelope. (a) A simplified sketch of a triaxial


experimental apparatus. A cylindrical sample placed in an impervious
rubber/copper jacket is given an axial load through a hydraulic ram and a
radial confining pressure through a fluid. (b) Stress-strain diagram for a
hypothetical experiment at 0.5 kbar confining pressure. (c) Mohr diagram for
the same hypothetical experiment. Each Mohr circle represents a discrete state
of stress during experiment. The largest Mohr circle represents state of stress at
fracture. (d) Mohr circle at fracture showing two possible orientations of shear
fractures (p and p’) oriented at ± 2θ to σn axis. (e) Mohr diagram showing Mohr
circles at fractures at different confining pressures. Mohr/fracture/failure
envelope is tangent to all the Mohr circles at two points except at very low
confining pressure. The Mohr envelope divides the Mohr diagram into stable
and unstable regions. (f) Different types fracture behaviour along the Mohr
envelope. See text for discussion.

53
The rock samples are cut into cylindrical shape, jacketed by weak and
impermeable layer of copper or rubber and put in a piston cylinder. The sample is
surrounded by a chamber filled with fluid through which confining pressure (σc) can
be varied and controlled. The rock sample is subjected to an axial load (σa) through a
piston driven by hydraulic ram. The axial load can be either σ1 or σ3 (σa = σ1 or σa = σ3).
The state of stress can, therefore, be either axial compression (σa = σ1 > σ2 = σ3 = σc) or
axial extension (σa = σ3 < σ2 = σ1 = σc). The triaxial testing apparatus can be modified
for experiments under uniaxial tension, but such experiments are rare. Data obtained
from such experiments are usually graphed in Mohr diagrams, and stress-strain
diagrams with stress and strain plotted on vertical and horizontal axes, respectively.
Let us consider a series of experiments on a dry homogeneous rock in a triaxial rig
and demonstrate the results of the experiments in a stress-strain diagram as well as in a
Mohr diagram (Fig. 7.4, Suppe 1985). The experiments are conducted under axial
compression such that σa = σ1 > σc. We first increase both σ1 and σc together until both
of them reach a value of 0.5 kbar and the rock sample acquires an initial natural strain
(ε0) parallel to the axis of the cylindrical sample, i.e., parallel to σ1. Note that the
difference between axial load and confining pressure is zero at this stage (i.e., σ1 - σc =
0). We now keep increasing the axial stress keeping the confining pressure constant.
The strains and Mohr circles at three discreet axial stresses (σ11, σ12 and σ13) are shown
in Figs. 7.4a and 7.4b, respectively. If at anytime during this stage of deformation the
axial stress is returned to 0.5 kbar, the axial strain returns to ε0. Eventually, the rock
sample fails at an axial stress of σ14. Let us assume that this stress is 4.5 kbar. We take
the sample out of the rig and find that one shear fracture has developed oriented at 24°
to the axial stress direction, i.e., the axis of the cylinder. If we repeat the experiment
with another sample of the same rock we get a fracture oriented at the same angle but
on the other side of the axis, i.e., together they form a conjugate pair. However, we can
not predict which of the two possible fractures will yield during a particular
experiment. We say that the fracture strength of the rock is 4.5 kbar at a confining
pressure of 0.5 kbar. The largest Mohr circle in Fig. 7.4c represents the state of stress at
fracture. We now keep only the Mohr circle at fracture and remove the all other Mohr

54
circles for convenience (Fig. 7.4d). The orientations of the two fracture planes (p and p')
are also shown in Fig. 7.4d. We now conduct similar fracture experiments at increasing
confining pressures and plot Mohr circles at fracture at each confining pressure in the
same diagram (Fig. 7.4.e). For each Mohr circle there will be a pair of fracture planes
represented by p and p'. Loci of these points give us a curve called Mohr, or fracture or
failure envelope, which divides the Mohr diagram into stable and unstable parts (Fig.
7.4e). The state of stress in the stable part does not lead to any kind of fracturing. The
state of stress in the unstable part leads to catastrophic fracturing of rock. The Mohr
envelope shows three types of brittle fracture behaviour (Fig.7.4f):

• Tensile fractures: Tensile fracture behaviour is shown at the point where the
Mohr envelope crosses the negative side of the σn axis. The Mohr envelope is
tangent to the Mohr circle only at this point so that, only one set of fractures
oriented perpendicular to maximum tensile stress can develop. The tensile
strength, T0, is independent of other principal stresses and varies between 50
and 200 bars for most common rocks. Note that T0, is the value at the point
where Mohr envelope intersects σn axis.
• Coulomb fracture behaviour: The fracture behaviour of rocks in the straight line
segment of the Mohr envelope (continuous line in Fig.7.4f) is called the
Coulomb fracture behaviour. Obviously, fracture strength increases linearly
with confining pressure in this part of the Mohr envelope. This behaviour is
common for many rocks at intermediate confining pressure with σ1
approximately greater than |5T0|. In order to describe the Coulomb fracture
behaviour, we can write a linear equation of the form:

|σs*| = C + σn tanφ (7.1)


|σs*| = C + µ σn (7.2)

where, σs* is the critical shear stress, i.e., the shear stress at fracture; c and µ are
the intercept and slope, respectively; and φ is the slope angle of the line. The
above equation is commonly called Coulomb fracture criterion. The constants C
55
and µ describes the failure properties of rocks. C is the cohesive shear strength,
which represents resistance to shear fracture on a plane with zero normal stress,
and µ and φ are called coefficient of internal friction and angle of internal
friction, respectively. Whenever, the state of stress on a plane of some
orientation satisfies equations (7.1) or (7.2), a shear fracture may develop on that
plane. For a triaxial stress, the criterion is satisfied on two planes where Mohr
circle is tangent to the Mohr envelope. These two planes are symmetrically
oriented about σ1 (or σ3) axis and intersect along σ2 axis. They form a set of
conjugate shear planes. Shear fracture may develop along any one of these two
planes but the fracture criterion does not predict which of the two orientations
will be preferred at failure. The θ and φ are related by the relation

θ = (90° + φ)/2 (7.3)

The Coulomb fracture criterion may also be expressed in terms of principal


stresses, such as:

σ1 = S + Kσ3 (7.4)

where, S and K are compressive strength and earth-pressure coefficient and given by

S = 2 C [√(µ2 + 1) + µ] (7.5)
K = [√(µ2 + 1) + µ]2 (7.6)

The fracture criterion in the form of equation (7.4) is plotted on a graph of


maximum vs. minimum principal stresses although Mohr circle does not plot
on such a graph.
• Transitional tensile behaviour: This behaviour is shown by rocks with state of
critical stress lying in between Coulomb fracture behaviour and tensile fracture
behaviour on the Mohr envelope. This behaviour is shown above a critical value
of tensile stress σ1, which is about three times T0. In this segment, the fracture
strength increases nonlinearly with increasing confining pressure and equation

56
(7.3) is invalid. Fractures are oriented in the range 0 to 30° to the maximum
compression. Joints probably form in state of stress characteristic of tensile and
transitional tensile segment.

Although fracturing of rocks in nature is a complex process, the above discussion


should be considered as a predictive tool for brittle fracturing. There are several factors
that control fracture formation in rocks, such as confining pressure (σc) and differential
stress (σ1 - σc), as we have already seen. Among other factors, fluid pressure plays a
major role in significant modification of Mohr circles. Fluid is an important agent in
many geological processes including fracturing. We discuss the effect of fluid pressure
(Pf) on Mohr circle in some detail (Fig. 7.5).

Unstable
Mohr envelope
+σs Unstable
Mohr envelope

+σs Stable Stable

Pf
Pf 2θ
T0
−σn σ3e σ1e σ3 σ1 +σn −σn σ3e σ3 σ1e σ1 +σn
State of
State of
applied 2θ
effective stress
stress

p'
State of State of
−σs −σs
effective
stress
applied
stress

(a) (b)

Figure 7.5. Effect of pore fluid pressure on Mohr diagram. (a) Low differential
stress leads to tensile fractures. (b) Relatively large differential stress leads to
shear fractures.

Water incorporated in the intragranular spaces during sedimentation form pore


fluids. Dehydration reactions during diagenesis and metamorphism release H2O-rich
fluid into pore spaces. If all the pore spaces are interconnected all the way to the
surface, we can calculate fluid pressure (Pf) at a given depth using a an equation we
learnt in high-school physics, viz., P = ρ g h, where P, ρ, g and h are pressure, density,
acceleration due to gravity and depth (i.e., height), respectively. Similarly, we can
calculate lithostatic pressure or vertical normal stress, σnL. Taking densities of water
(ρw) and rock (ρr) as 1.0 and 2.3 gm/cm3, respectively, we get:

57
ω = Pf /σnL = (ρw g h)/(ρr g h) = 0.4 (7.7)

So it seems that the pore fluid pressure should be about 40% of lithostatic
pressure. However, it does not always work that way. In many deep drill wells in
sedimentary basins, the ratio ω approaches unity. In metamorphic petrology this ratio
is one unless otherwise specified. It is unreasonable to assume constant density for
easily compressible fluid although the same assumption may not be too bad for rather
incompressible rocks occurring in the upper part of the crust. The value of ω should
increase, as density of fluid increases with depth. It is for nothing that a metamorphic
petrologist will argue that water and fluids are not exactly same things! Secondly, pore
spaces may be not interconnected all the way to the surface. There may be
impermeable barriers preventing fluids at depth communicating with fluids at surface
and building up fluid pressure (formation overpressure).
The effect of fluid pressure on fracturing of rocks can be best explained through the
concept of effective stress. When a component of stress is lowered by an amount equal
to the pore fluid pressure, it is called effective stress. Thus,

σne = σn - Pf
σ1e = σ1 - Pf
σ2e = σ2 - Pf (7.8)
σ3e = σ3 - Pf
σce = σc - Pf
σemean = (σ1 + σ2 + σ3 - 3Pf)/3

where superscript e denotes effective stress. Shear stresses are not affected because
fluid at rest cannot sustain shear stress.
The size of the Mohr circle remains between state of effective stress and state of
stress with zero fluid pressure. This is because differential stresses in both cases are
same, i.e., (σ1 - σ3) = (σ1e - σ3e). The only effect is seen in translating of the Mohr circle
towards lower compressive stress (i.e., towards left) in the Mohr diagram by an

58
amount equal to Pf (Fig. 7.3). The equations describing fracture criterion also remain
same except that σne replaces σn:

|σs*| = C + σne tanφ = C + (σn – Pf) tanφ (7.9)


σ1e = S + K σ3e, or (σ1 - Pf) = S + K (σ3 - Pf) (7.10)

Shifting of Mohr circle to lower compressive stress may lead to hydraulic fracturing
via following processes:

• If the differential stress [(σ1 - σ3) or (σ1e - σ3e)], which represents the size of the
Mohr circle, is small and σ3e = T0, then extension fractures can form (Fig. 7.3a).
• If the differential stress is relatively large, the Pf may drive the Mohr circle far to
the left to touch the fracture envelope and shear fracturing may occur (fig. 7.3b).

Putting it together we may state that given the initial state of stress in the rock is
within the stable part of the Mohr diagram, the fluid pressure may be increased to a
level that the Mohr circle for state of effective stress may touch the fracture envelope
for fracturing to occur. In order to increase the permeability of hydrocarbon reservoir
rocks, artificial hydraulic fracturing is carried out by pumping fluids through oil wells.
Strength and ductility of rocks increase with mean stress. The pore fluid pressure
reduces mean stress and makes rocks weaker and more brittle. Therefore, high fluid
pressure may lead to brittle fracturing at depths normally associated with ductile
deformation.

59
60
8. Faults
A fault is a brittle shear fracture along which appreciable amount of differential
displacement has taken place. This rather straightforward and generally accepted
definition of faults has one serious problem, i.e., with the word appreciable. The
displacements along faults may vary from microscopic to hundreds or thousands of km
on very large faults, such as transcurrent faults. Therefore, what constitutes a
displacement to be appreciable depends on the scale of observation. For example, a
geologist working on regional faults pattern may dismiss a fault seen at an outcrop
with centimeter/millimeter-scale displacement as minor shear fractures or microfaults. A
fault with few hundred meters of displacement may lead to the formation of a good
trap for hydrocarbon accumulation but may not show up on a regional-scale seismic
survey.
A fault may occur as one sharp plane of discontinuity (Fig. 8.1a) or as a zone of
closely spaced planes of discontinuity (Fig. 8.1b). The discontinuity surfaces are the
brittle shear fractures. Faults at depth may not have such sharp discontinuities; instead
the deformation may be distributed in a diffused zone. Such zones of distributed
deformation are called ductile shear zone (Fig. 8.1c). The maximum strain in a ductile
shear zone is at the middle of the zone and gradually decreases to zero at the shear
zone wall. In some cases a fault zone may have components of both brittle and ductile
deformations (Fig. 8.1d).

(a) (b) (c) (d)

Figure 8.1. Styles of faulting. (a) Brittle fault on a single surface. (b) Brittle fault
zone consisting of several subsidiary fault surfaces. (c) Ductile shear zone. (d)
Brittle-ductile shear zone.
61
8.1 Fault terminology

The terminology associated with faults is complex and ambiguous. In this section
and in the sections to follow, the terms used are most prevalent, have precise meaning
and most useful in describing faults (see Fig. 8.2):

• Blind/emergent: If tip line does not reach the ground surface the fault is blind.
When the tip line is exposed to the ground surface it is an emergent fault. A
blind fault may become emergent through erosion.

emergent fault blind fault


compressional decollement
tip
fault trace
tip line
tip line
fault
surface

extensional decollement

hangingwall emergent fault


cut-off point

tip
blind fault

compressional detachment
footwall
cut-off point listric fault

hangingwall hangingwall
footwall cut-off line block
cut-off line
footwall
block

extensional detachment

N N : Net slip
D: Dip slip
D S S : Strike-slip
V component
H : Horizontal
H component
V : Vertical
component

Splay faults
isolated diverging
splay splay rejoining
splay connecting
splay

b.p.
b.l.
t.l.
b.p.
b.l.
b.l.

b.p. - branch point


b.l. - branch line
t.l. - tip line m ain
fa ult

Figure 8.2. Diagrams illustrating different geometric terms associated


with faults.

62
• Branch line: The line of intersection between any two fault surfaces is known as
the branch line. If the faults crop out at the surface, this line will appear as the
branch point. In sections, they appear as branch point.
• Cut-off line: The line along which a marker horizon is truncated by a fault. For
the same marker horizon there are two cut-off lines on two sides of a fault (e.g.
hangingwall cut-off line and footwall cut-off line for inclined faults). In sections,
they appear as cut-off point.
• Décollement: A low-angle regional fault (compressional or extensional), which is
usually parallel to a soft or incompetent strata.
• Detachment fault: A low-angle regional fault (compressional or extensional) that
cuts through gently-dipping or horizontal strata.
• Fault trace: The line along which the fault plane cuts the ground surface is
known as fault trace. Blind faults do not have fault trace!
• Hangingwall/footwall: For an inclined fault, the block above the fault surface is
known as hangingwall and the block below the fault plane is known as
footwall. In a tunnel parallel to the fault surface, the hangingwall literally hangs
overhead and the footwall lie under the foot of an observer. For a fault with
vertical dip, there is no hangingwall or footwall.
• Horse: A mass of rock surrounded on all sides by fault.
• Listric fault: A fault with smoothly curving surface is called a listric fault.
• Separation: The distance between points on cut-off lines on either side of fault.
The separation distance varies depending on the direction in which it is
measured. The horizontal separation is the separation in a horizontal plane and
may or may not equal the strike-slip component.
• Slip on a fault: The distance between two points on the fault plane those were
together before faulting, represents total displacement or net slip. The net slip is
a vector quantity. The net slip can be resolved into strike-slip and dip-slip
components. The dip-slip component can further be resolved into a horizontal
component and a vertical component.

63
• Splay: A secondary fault (i.e., smaller in size and displacement) that emerges
from a main fault. There are different types of splays. A connecting splay
connects two faults whereas a rejoining splay rejoins the same fault from which it
branched off. When a splay fault branches off and diverges away from a fault, it
is called diverging splay. From an isolated splay no other fault branches off.
• Tip line: Individual faults are always of limited extent and displacement die
away to zero. The line along which fault displacement becomes zero is known
as tip line (tip or tip point in two dimensions). Tip line may emerge on the
erosion surface as tip point or it may remain blind.

8.2 Fault classification

Faults are classified on the basis of relative movement between the walls. This
scheme of classification is adequate under most situations (Fig. 8.3).

• Dip-slip faults: These faults have dominant dip-slip component. The strike-slip
component is small or negligible. They are of two types: (1) Dip-slip faults in
which hangingwall moves down relative to the footwall are called normal fault.
The dip of normal faults is usually 50° or more. They are sometimes called
extensional fault, as they are more common in extensional tectonic set up, such as
divergent plate margins. (2) Reverse faults are those in which hangingwall moves
up relative to footwall block. A reverse fault in which dip of the fault surface is
less than 45° are called thrust fault. They are sometimes called compressional fault,
as they are more common in compressional tectonic set up, such as convergent
plate margins.
• Strike-slip faults: Strike-slip faults are those in which strike-slip component is
dominant, i.e., predominantly horizontal displacement. They are steeply
inclined, often vertical. They are characteristic of translational tectonic set up.
Large-scale sub-vertical strike-slip faults are sometimes called transcurrent or
wrench fault. They are of two types: sinistral (syn. left-lateral or left-handed)

64
when the relative movement is counterclockwise or dextral (syn. right-lateral or
right-handed) when the relative movement is clockwise.
• Oblique-slip faults: In these faults the strike-slip and dip-slip components are
approximately same. They may be right normal, right reverse, left normal and
right reverse depending on whether the opposite walls move clockwise or
counterclockwise and whether hangingwall moves up or down relative to
footwall.

(a) Dip-slip fault (net slip = dip slip) (b) Strike-slip fault (net slip = strike slip)

Normal Reverse

Sinistral Dextral

(c) Oblique-slip fault (dip slip strike slip) (d) Rotational fault

Right normal Right reverse Left normal Left reverse

Figure 8.3. Classification of faults based on relative displacements between the


walls.

The normal, reverse and strike-slip faults are the most common types of faults.
They typically form in extensional, compressional and translation tectonic regimes,
respectively. The terms strike fault, dip fault, transverse fault, wrench fault, tear fault,
transcurrent fault, hade, throw, and heave should not be used because some of them
are obsolete while others are imprecise and may create confusion.

8.3 Anderson's theory of faulting

There is no single comprehensive theory of mechanics of faulting. The theory of


faulting as presented by E. M. Anderson (1951) is the simplest and the most
enlightening (Fig. 8.4). There are two assumptions involved in the Anderson's theory of
faulting:

65
• Shallow level faults are Coulomb fractures. A consequence of this assumption is
that there are two possible orientations of Coulomb fractures with respect to the
principal stress directions. The line of intersection of two fractures is parallel to
intermediate principal stress axis, σ2. The direction of maximum principal
compression, σ1, bisects the acute angle between the fractures and the obtuse
bisector is parallel to the least principal compression direction, σ3. The material
shortens parallel to σ1 and extends parallel to σ3 as a result of slip along the
fractures.
• In shallow-level faulting one of the principal stress directions is vertical. This is due to
the fact that the earth-air interface is a plane of no shear. Vertical σ1, σ2 and σ3
leads to normal, strike-slip and reverse faults, respectively. These are the three
most common types of faults, as noted earlier.

σ1 σ3 σ2

σ3 σ1 σ3 σ1
σ2
σ2

U D D U

σ3 σ1
σ1 σ3 σ1
σ2

(a) (b) σ2 (c)


σ2 σ3

Figure 8.4. Anderson’s theory of faulting. Vertical σ1, σ3 and σ2 lead to normal
faulting (a), thrust faulting (b) and strike-slip faulting (c), respectively. D and U
are downthrown and upthrown sides.

The Anderson's theory of faulting does not predict some of the commonly observed
features associated with faulting, such as, high-angle reverse faults, very low-dipping
and very high-dipping normal faults, listric faults and transform faults. Yet, it remains
a simple and elegant way of describing how faults form in different tectonic settings.

66
8.4 Recognition of faults

Under the present-day scenario of doing everything on a computer monitor it has


become imperative to emphasize an obvious fact, viz., faults must be recognized on
geological criteria. Several categories of geological features are used to recognize faults.
They are not mutually exclusive; often more than one category of features are found
associated with a single fault.

• Geomorphic features: The most obvious geomorphic feature that may betray the
presence of a fault is fault scarp (Fig. 8.5). Fault scarps are continuous linear
features characterized by sudden break in topographic slopes. They may
indicate either active or inactive faults. However, the original scarp associated
with an old and inactive fault may not survive erosion for very long. Regular
fault scarps are more commonly associated with normal faults (Fig. 8.5a).
Subsidiary smaller scarps are usually present parallel to the main scarp. The
uplifted block is usually cut by V-shaped side valleys. The erosional debris
derived from the uplifted block and brought along the side valleys form alluvial
fans on the downthrown block. Successive movements along faults may leave
perched alluvial terraces on the side valleys. The scarps formed due to thrust
faulting tend to be irregular (Fig. 8.5b). An interesting feature associated with
thrusting is that the fault often overrides the debris derived from the uplifted
hangingwall and deposited in front of the scarp. In strike-slip faulting the scarp
is usually small and of local importance (Fig. 8.5b). Deflected or offset
geomorphic features, such as, river channels, hogbacks and ridges may indicate
the presence of strike-slip faulting. The scarps associated with large faults may
show up as lineament satellite images and air photographs. But it must be
remembered that a majority of lineaments drawn on satellite images are not
faults. Unfortunately, there is a rising but very unscientific tendency in some
quarters to draw lineaments on satellite images, construct a rose diagram and
deduce stress axes. Surface exposures of faults must be sought out during

67
fieldwork because much critical information can be gathered through field
studies of faults.

Normal fault Strike-slip fault


Thrust fault Deflected river channels
Main fault scarp Perched terrace Offset hogbacks
Subsidiary fault scarp Local scarp
Irregular scarp
Alluvial fan Erosion debris

(a) (b) (c)


Main fault
Subsidiary fault

Figure 8.5. Geomorphic features associated with three main types of faulting.

2 2 km 1
F1
N 2
1 F2 F3
3
F1
1
2

3 2
3
4 F2
7
3

2
8
1

2 9
(a) (b)

Figure 8.6. Recognition of faults from map patterns. (a) Offsets of rock units
trace faults. Symmetric repetition of rock units in the N-S direction around
rocks 1 and 4 are due to folding. The fault surfaces trace lines of discontinuities
along which three rock units meet. (b) Omission and repetition of rock units
trace faults F1 and F2.

• Geological map and stratigraphy: Large faults are relatively easy to recognize in
regions of moderate to excellent exposures of rocks through systematic
mapping. One should be careful, however, because some of the geologic
features are common to both faults and unconformities. Faults are recognized
on the basis of truncation and offset of one or more rock units (Fig. 8.6a).

68
Truncation and offset occur along a line on a map that defines a discontinuity
along which three rocks meet at a point. Such a map pattern may also indicate
angular unconformity if rocks on two sides of the discontinuity are of different
ages. In case of faults, same rock units should be present on both sides of the
discontinuity. This relation may not be valid for large overthrusts, which may
bring older rocks to lie over younger rocks or may even bring metamorphic
rocks on top of sedimentary rocks. The map patterns due to truncation and
offset vary considerably depending on the orientations of faults relative to the
orientations of rock layers affected by faulting, and amount and direction of
slip. The map patterns may become even more complex if the terrain had an
earlier history of folding and faulting. A common effect of truncation and offset
is an apparent horizontal shifting of the rock units. There may not be any
truncation and offset if the strike of the fault is same as the strike of the rock
units. In such cases, faults can be recognized on the basis of repetition and
omission of strata if the stratigraphy of the rock units is known (Fig. 8.6b). The
line along which a packet of rock units are repeated (fault F1 in Fig. 8.6b) or
some ofthe rock units are omitted (fault F2 in Fig. 8.6b) marks a fault plane.
Symmetric repetition of rock units about a particular rock unit due to folding
(see Fig. 8.6a) should not be confused with simple repetition due to faulting. In
drill wells, missing or repetition of beds can be used to predict faults.
• Fault rocks: Large-scale faulting with significant amount of displacement lead to
development of characteristic textures and structures within the rocks present
in the fault zone. Rocks with such characteristic textures and structures are
collectively called fault rocks (Wise et al. 1984), which can be broadly divided
into cataclasite and mylonite. Cataclasites or cataclastic rocks is a general term
that refers to rocks fractured into clasts or ground into powder and signify
brittle deformation. Individual clasts are sharp, angular and internally
deformed. Cataclasites usually do not have any planar and linear fabrics. Fault
gouge and fault breccia are results of cataclastic deformation within fault zones.
Incohesive (i.e., friable) cataclasitic rocks are characteristic of faulting above

69
depths of 1-4 km; below this depth cataclastic rocks are cohesive. Frictional
heating during brittle faulting can be sufficient high to melt a small portion of
the rock. The melt may intrude into surrounding fractures and quenched to
give veins of pseudotachylite. Mylonitic rocks form if faulting occurs at depths
exceeding 10 to 15 km. These are fine-grained rocks with grain size reduction
via dynamic recrystallization and neomineralization. Mortar texture with
highly strained clasts in a matrix of fine-grained recrystallized grains is a typical
texture in these rocks. Planar and linear fabrics are common in mylonites. This
texture is a characteristic feature of ductile shear zones. The development of
textures and structures in fault zones depend on several parameters including
strain, strain rate, temperature, pressure and pore fluid pressure. Therefore, it is
not possible to make simple correlation between depth of faulting and type of
textures and structures in fault rocks.

8.5 Separation and displacement

Marker beds or quartz/calcite veins may get displaced during faulting and provide
unambiguous evidence of faulting. Fault separation distance is the distance measured
between hangingwall and footwall cut-off lines of the same horizon. The separation
distance varies depending on the direction along which it is measured. Out of infinite
number of separation distances, only the one measured in the direction of net slip will
give the true displacement. Obviously, fault separation and displacement are separate
parameters. Care must be taken to interpret offset marker layers because the geometry
of offset depends on the orientation of the layers prior to faulting, in addition to the
orientation of the displacement vector. For example, the marker beds will not show any
offset if the displacement direction and cut-off lines are parallel. For this reason,
horizontal beds affected by strike-slip faulting do not show any offset.
In Fig. 8.7a, a set of inclined beds are affected by strike-slip faulting. The marker
layers show offset on both horizontal and vertical surfaces (Fig. 8.7b). Horizontal and
vertical offsets give horizontal separation and vertical separations, respectively. Note that
they are not necessarily equivalent to strike-slip and dip-slip components. So, what

70
appears on a cross section (or on a seismic reflection profile) to be a normal fault
displacing sub-horizontal beds may actually be strike-slip fault affecting inclined beds.
Fig. 8.7c shows a set of inclined beds affected by normal faulting. If the fault scarp is
completely eroded away to give a flat topography, the fault will appear like a strike-
slip fault. Oblique-slip faults show offset on both map view and cross section,
suggesting strike-slip faulting and dip-slip, respectively. Therefore, three dimensional
reconstruction and presence of lineation may be required for unambiguous
interpretation of offset markers on map as well as on section.

strike-slip fault

(a) normal fault (b)

(c) (d)

Figure 8.7. Fault separation and displacement. (a,b) Inclined beds


affected by strike-slip faulting. Horizontal and vertical offsets suggest
strike-slip and normal faulting, respectively. (c,d) Inclined beds affected
by normal faulting. Offsets on horizontal surface indicate apparent
strike-slip faulting.

8.6 Determination of fault displacement

When we describe movement on a fault plane, we consider the movement of one


block while keeping the other block fixed owing to the fact that it is not possible to
have a reference frame independent of both the faulted blocks. For example, we define
a normal fault as the fault in which the hangingwall goes down relative to the footwall.
We could as well state that in a normal fault footwall goes up relative to the
hangingwall. Alternately we could also say that in normal fault both the blocks go up,
footwall goes up more than hangingwall. Finally, it is equally valid to state that both
the blocks go down, hangingwall goes down more than the footwall. Simply put, we
are unable to determine absolute movements on faults. The faulting that caused the
71
devastating 1964 Alaska earthquake near Anchorage is a rare example where absolute
movement of both the fault blocks with respect to an external reference frame (sea
level) could be determined (Platker 1965). Measurements at one place where normal
faulting had occurred showed that both the fault blocks had moved upward – one
block had moved more than the other. We can only hope to determine the orientation
of the displacement vector and the sense of shear movement in most outcrops of fault
or fault zone. The magnitude of the displacement can rarely be determined.
Differential displacements often lead to finely polished fault surfaces, called
slickenside surfaces. Ridge-in-groove or striation lineations commonly form on slickenside
surfaces in the direction of displacement. Growth of crystal fibers may result from fluid
flowing through small spaces opened up along fault surfaces. The crystal fibers trace
slickenfiber lineations, which also parallel displacement direction.
A large number of shear sense indicators have been proposed for both brittle and
ductile shear zones, some of which can be rather tenacious in most outcrops. A few
“better” indicators are discussed below:

• Drag fold: If the walls of a fault undergo ductile deformation, then marker layers
may be dragged into fold forms (Fig. 8.8). These folds are called drag folds
whose hinge lines are parallel to the cut-off line and oriented at high angles to
displacement direction. The sense of shear displacement is opposite to the
curvature of the fold. Drag folds do not form if the cut-off lines are at low angle
to the slip direction. Reverse drag associated with rollover anticlines have
opposite sense of displacement to that of drag folds.

(a) (b) (c)

Figure 8.8. (a,b) Drag folds giving sense of displacement. The beds are dragged
into the faults in the direction opposite to the sense of displacement. (c)
Opposite sense of displacement in reverse drag in rollover anticline.

72
• Steps associated with slickenfibers: Exposed slickenside surfaces with slickenfiber
lineation commonly have small steps (also called chatter marks) oriented
approximately normal to the lineation (Fig. 8.9a). All the steps on the slickeside
surface face in the same direction. The missing block on the outcrop (i.e., the
block eroded away) is thought to have moved in the direction opposite to the
stepped surface. The surface feels smooth to the hand if rubbed in the direction
of the movement of the missing block.

slickenfiber steps

(a) (b)

s-surface

c-surface

(c) (d)

median line

center line

(e) (f)

Figure 8.9. Examples of small-scale structural features as shear sense indicators.


(a) Slickenfiber lineation and steps. (b) En-echelon gash veins and asymmetric
folds. (c) S-C fabric. (d) Bookshelf gliding of pophyroclasts. (e) σ-structure. (f) δ-
structure.

• En-echelon gash veins and asymmetric folds: These features develop in ductile shear
zones. En-echelon gash veins are tension veins formed normal to maximum
stretching direction (Fig. 8.9b). Acute angles formed by gash veins and shear
zone walls typically point in the direction opposite to the direction of shear
displacement. The axial planes of asymmetric folds are oriented normal to the
73
maximum shortening direction. The sense of shear is in the direction of the
acute angle between axial plane and the shear zone wall.
• S-C fabric: Schistosity is well-developed in many ductile shear zones (Fig. 8.9c).
Owing to heterogeneous strain the schistosity surfaces (s-surfaces) are usually
sigmoidally curved. Within shear zones, zones of high and low strains parallel
the shear zone walls and give rise to a schistosity (c-surfaces) oriented parallel
to the shear zone walls. Overall the shear zone displays what is known as s-c
fabric. The acute angle between s and c surfaces point to the shear direction.
• Bookshelf sliding: If fractured clasts or porphyroclastic minerals such as feldspar
or micas are caught up in shear zones, the individual parts rotate in the
direction of shear resulting in bookshelf structure, which may be visible to
naked eyes. If the fractures initially make high angle to the shear plane, the
shear sense in the fractured clasts is opposite to that of in the matrix (Fig. 8.9d).
If the initial angle is low, the shear sense on the fractures is same as it is the
matrix.
• σ- and δ-structures: These structures are formed around porphyroclasts and are
best seen in oriented thin sections under microscope (Passchier and Simpson
1986). The shape of the asymmetric pressure shadow trails containing
dynamically crystallized grains defines these two types of structures. In σ-
structure (Fig. 8.9e), the median lines drawn through the pressure shadow do
not cross the center line drawn through the center of the clast and oriented
parallel to the average cleavage surface. In δ-structure (Fig. 8.9f), the median
lines cross center lines. The orientations of pressure shadows relative to sense of
shear are shown in Figs. 8.9e, f.

All the shear sense indicators in Fig. 8.9 are right-handed or dextral. If we look at
the same indicators from the backside of the page, they will all appear to left-handed or
sinistral. Therefore, it is not sufficient to state that the displacement is dextral or
sinistral. The movement of blocks with respect to geographic coordinates should also
be known for definitive interpretation.

74
9. Thrust Faults
Low angle (dip < 45°) reverse faults are frequently termed as thrusts or thrust faults.
Thrust faults are much more common than steeply dipping reverse faults. The thrust
faults are dip-slip faults with zero or negligible strike-slip component of displacement.
They are abundant in the upper crustal levels of the external zones of compressional
orogenic zones. Consequently, they are also referred to as contractional faults by some
workers. They bring older rocks to lie over younger rocks and often result in large-scale
vertical duplication of regionally sub-horizontal strata. Thrusts cause elevation of
hangingwall relative to footwall giving rise to irregular fault scarps. They range in scale
from millimeters to meters, through tens to hundreds of kilometers in fold-thrust belts
to thousands in kilometers in convergent plate margins.
Over the years a plethora of thrust fault-related terminologies have been proposed
in the literature, many of them with obscure meaning. Terminologies defined here are
mostly after McClay (1992) who made a brave attempt to bring in a sense of sanity in
an otherwise chaotic and often confusing set of terminologies littered through
literature.

9.1 Ramp-flat thrust geometry


Thrust surfaces usually have stair-case geometry (Fig. 9.1). Though the thrust
trajectories shown in Fig. 9.1 are idealized, many thrust have approximately stair-case
trajectory. A thrust surface may be considered to have two sides, a hangingwall side
and a footwall side. The following terms are used to describe flat-ramp geometry of
thrust faults:
• Flat: Sub-horizontal or gently-dipping (at the time of initiation) part of a thrust.
Flats usually propagate through weaker layers.

75
Fault surface Cut-off line
Flat Ramp Flat Ramp Flat
RamFlat
Flat p
Flat Ram
p

(a) Cut-off point (b)


Hangingwall side
of the thrust

Tectonic transport
direction
HWR
HWR HWF Oblique
ramp
HWF Lateral
ramp

Frontal
ramp
FWF FWR FWF FWR FWF
(c) Footwall side
(d)
of the thrust

Figure 9.1. (a,b) Flat-ramp-flat (stair-case) thrust trajectory in 2-D and 3-D. (c)
Hangingwall and footwall blocks separated to illustrate ramps and flats in the two
blocks. FWF: footwall flat, FWR: footwall ramp, HWF: hangingwall flat, HWR:
hangingwall ramp. (d) Ramp in 3-D showing frontal, oblique and lateral ramps.

• Ramp: The moderately-dipping part (at the time of initiation) of a thrust. Ramps
usually climb up-section across stiffer layers.
• Hangingwall flat (HWF): The portion of the thrust where the fault is parallel to
the bedding surfaces on the hangingwall side.
• Hangingwall ramp (HWR): The portion of the thrust where the fault is oblique to
the bedding surfaces on the hangingwall side.
• Footwall flat (FWF): The portion of the thrust where the fault is parallel to the
bedding surfaces on the footwall side.
• Footwall ramp (FWR): The portion of the thrust where the fault is oblique to the
bedding surfaces on the footwall side.
• Frontal ramp: The strike of the ramp is perpendicular to the regional tectonic
transport direction.
• Lateral ramp: The strike of the ramp is parallel to the regional tectonic transport
direction.
• Oblique ramp: The strike of the ramp is oblique to the regional tectonic transport
direction.

76
• Thrust trajectory: The trace of a thrust on a cross section. Thrust faults always cut
up-section in undeformed rock but may cut down section in previously folded
terrains.

Flats and ramps on thrust trajectory, and hangingwall footwall flats and ramps are
different aspects of flat-ramp geometry.

9.2 Thrust vergence

Thrust vergence (Fig. 9.2) refers to the direction towards which hangingwall moves
relative to footwall. Thrust faults are most common in fold-thrust belts (FTBs) typical of
contractional orogenic setting. The undeformed sedimentary basin in front of an FTB is
called foreland and the internal part of an FTB is known as hinterland, which may
contain metamorphic rocks involved in ductile deformation. (Fig. 9.2):

• Hinterland-vergent thrust: The hangingwall of the thrust fault moves towards the
hinterland.
• Hinterland-dipping thrust: The dip of the thrust is towards the hinterland. This
type of thrust is also called forethrust because the vergence is towards the
foreland.
• Foreland-vergent thrust: The hangingwall of the thrust fault moves towards the
foreland.

Hinterland Ramp anticline Ramp anticline Foreland


Foreland vergent
Hinterland vergent

(a) Hinterland dipping (forethrust) Foreland dipping (backthrust)

(b)

Figure 9.2. (a) Thrust vergence. (b) Pop-up structure.

77
• Foreland-dipping thrust: The dip of the thrust is towards the foreland. This type
of thrust is also called backthrust because the vergence is towards the
hinterland.
• Pop-up: The portion of hangingwall block that has been uplifted by a
combination of forethrust and backthrust (Fig. 9.2b).

9.3 Thrust sheet (Fig. 9.3)

• Thrust sheet: The areal extent of the hangingwall block of a regionally important
low-angle thrust is commonly much greater than the thickness. Such a tabular-
shaped hangingwall block is called a thrust sheet. Thrust sheets are given the
same name as the underlying thrusts, such as Jutogh thrust sheet and Jutogh
thrust in Himachal Himalayas.
• Thrust nappe: The French word nappe means a sheet and, therefore, thrust sheet
and thrust nappe should be synonymous. However, the term thrust nappe is
reserved for thrust sheets with significant movement (>10 km) relative to
footwall.
• Allochthon: An adjective used to describe a thrust sheet that has moved large
distance from its original position. The rocks within an allochthon are thus
geologically out of place and are called allochthonous. Obviously, allochthon and
nappe are closely related terms.
• Parautochthon: An adjective used describe a thrust sheet that has smaller relative
displacement as compared to allochthon.
• Autochthon: A large region of rock that has not moved from its original position
is called autochthon. The rocks within an autochthon are called autochthonous.
The basement rocks underlying a thrust are autochthonous.
• Overthrust: A large thrust in which the hangingwall (i.e., the thrust) has actually
moved relative to footwall.
• Underthrust: A large thrust in which the footwall (i.e., the thrust) has actually
moved relative to hangingwall.

78
• Klippe: The erosional remnant of a once continuous thrust sheet above a
horizontal or gently-dipping thrust is known as klippe.
• Window: If erosional processes have penetrated to deeper levels of a thrust
sheet, then outcrops of footwall can be exposed. A closed outcrop of footwall
rocks framed by overlying thrust sheet is called window.
• Root zone: The region in the hinterland direction where a thrust sheet passes into
the subsurface is known as root zone. No sense of “origin” or ‘genetic overtones”
should be attached to this term.

Allochthonous rocks
Klippe

Window

5 km
Root zone Autochthonous rocks

Figure 9.3. Cross section of thrust sheet showing allochthon and autochthon as
well as window and klippe.

9.4 Ramp anticline

In association with ramp-flat thrust trajectory, anticlines form in the hangingwall


just above the ramps. Such anticlines are called ramp anticline or hangingwall anticline
(Fig. 9.4). They form due to movement of the thrust sheet up and over the ramp.
Different components of ramp anticlines are as follows:

• Leading anticline/syncline pair: If the ramp anticline is flat crested, the anticline
and syncline pair located towards the transport direction of the ramp anticline.
• Trailing anticline/syncline pair: If the ramp anticline is flat crested, the anticline
and syncline pair located in the direction opposite to the transport direction of
the ramp anticline.
• Forelimb: The fold limb of the ramp anticline located towards the transport
direction is termed as forelimb.
• Backlimb: The fold limb of the ramp anticline located opposite to the forelimb is
termed as backlimb.

79
• Regional: It is the original elevation of a particular stratigraphic unit or datum
surface, measured at a place where it is not involved in any thrust related
structure.
• Leading edge: The edge of a thrust sheet towards the transport direction.
• Trailing edge: The edge of a thrust sheet in the direction opposite to the transport
direction.

Hinterland Transport Foreland


direction
trailing leading axial
anticline anticline surface

trailing leading
syncline ramp syncline ramp
anticline forelimb anticline
backlimb leading
trailing edge
edge

Regional

Figure 9.4. Diagram illustrating different components of ramp anticlines.

9.5 Thrust systems

Thrust systems are made up of a group of link thrusts that are geometrically,
kinematically and mechanically related. Majority of the thrust systems can be grouped
into either duplexes or imbricate thrust systems. A duplex is an array of thrusts linking a
floor thrust (sole thrust) at the base to a roof thrust at the top (Fig. 9.5a). A closely related
array of thrusts all of which merge into a floor thrust is called imbricate thrust system
such that thrust sheets overlap like roof tiles (also called imbricate fan or schuppen
structure) (Fig. 9.5b). There is no roof thrust in this thrust system. It may be difficult to
distinguish between an imbricate fan and a duplex whose roof thrust has been eroded
(eroded duplex). A third type of thrust system, called triangle zone, is also fairly common
in fold thrust belt.

80
Duplex Roof thrust
Imbricate fan

Floor thrust
(a) Link thrusts (b)

Figure 9.5. Two main types of thrust systems – duplex (a) and imbricate fan (b).
Note that in imbricate fan there is no roof thrust.

Duplexes

Geometry of a duplex depends on ramp angle, ramp height, initial thrust spacing
and displacement on individual thrusts. By varying these parameters, a bewildering
variety of duplex geometry can be obtained. A three-fold classification of duplexes is
shown in Fig. 9.6 (McClay 1992, modified after Mitra 1986): (1) independent ramp
anticlines and hinterland-dipping duplexes, (2) true duplexes including ramp anticline
footwall, ramp anticline hangingwall and frontal zone of ramp anticline, and (3)
overlapping ramp anticline including antiformal stack and foreland-dipping duplex.
The different types of duplexes are as follows (Figs. 9.6, 9.7):

• Independent ramp anticline: The final thrust spacing is much larger than
displacement on individual thrusts leading to formation of widely spaced ramp
anticlines. Ramp anticlines do not interfere with each other.
• Overlapping ramp anticlines: The horses stack up on top of each other in such a
way that the overall geometry is like an antiformal arch.
• Antiformal stack: This is a variation of overlapping ramp anticlines in which the
overlapping horses have coincident trailing branch lines.
• Breached duplex: A duplex in which out-of-sequence movement on link thrusts
have breached or cut through the roof thrust.
• Corrugated or bumpy roof thrust: A duplex in which the roof thrust is corrugated.
• Hinterland-dipping duplex: Both link thrusts and bedding dip towards hinterland.
With increased displacement, independent ramp anticlines grade into
hinterland-dipping duplex.

81
• Foreland-dipping duplex: At least part of link thrusts and bedding within the
duplex dip towards foreland. With increased displacement, hinterland-dipping
duplex may grade into foreland-dipping duplex.
• Frontal zone of ramp anticline: The duplex is at the front of a ramp anticline.
• Passive roof duplex: A duplex in which sequence above the roof thrust has not
been displaced towards the foreland.

Independent ramp anticline Hinterland-slopping duplexes

Increased
displacement

Ramp anticline footwall

True duplex Ramp anticline hangingwall

Second
order
duplexes

Frontal zone of ramp anticline

Overlapping ramp anticlines


leading to antiformal stack
Foreland-dipping duplex

Increased displacement

Figure 9.6. Classification of duplexes (after Mitra 1986, modified by McClay


1992).

• Planar roof duplex: A duplex in which top of the roof thrust is planar.
• Ramp anticline footwall: Duplex is in the footwall of the ramp anticline.
• Ramp anticline hangingwall: Duplex is in the hangingwall of the ramp anticline.
• Truncated duplex: A duplex that is truncated (or beheaded!) by an out-of-
sequence thrust.

82
• True duplex: Parts of all the link thrusts and roof thrust are parallel to the frontal
ramp.

Roof thrust

Floor thrust
(a) Planar roof duplex

Corrugated
roof thrust
(b) Antiformal stack

Breaching
thrusts

(c) Corrugated or bumpy roof duplex

(d) Breached duplex

Passive roof Tip of buried


thrust thrust

(e) Truncated duplex


(f) Passive roof duplex

Figure 9.7. Some examples of variation in duplex geometry.

Imbricate thrust systems

• Leading imbricate fan: Maximum displacement is on the leading (i.e., lowermost)


thrust.
• Trailing imbricate fan: Maximum displacement is on the trailing (i.e., highest)
thrust.
• Blind imbricate fan: All the thrusts are buried below the erosion surface. Folding
at higher structural level compensates the displacement along buried thrusts.

Erosion level

(a) Leading imbricate fan (b) Trailing imbricate fan (c) Blind imbricate fan

Figure 9.8. Types of imbricate fans. (a) Leading imbricate fan with maximum
slip on the frontal most thrust. (b) Trailing imbricate fan with maximum slip on

83
the trailing thrust. (c) A blind imbricate fan with folding as the surface
manifestation.
Triangle zones

A combination of two thrusts with the same basal detachment but with
opposing vergence forms a triangular zone (Fig. 9.9a). A pop-up structure (Fig. 9.2b) is
also type of triangular zone. An intercutaneous thrust wedge (Fig. 9.9b) is bound by a
floor thrust at the base and a passive roof thrust at the top.

(a) Triangle zone (b) Intercutaneous wedge

Figure 9.9. Triangle zone (a) and intercutaneous wedge (b).

Evolution of hinterland-dipping duplex

Kinematic modelling of duplexes and imbricate fans requires that the values of
several parameters, such as ramp angle, ramp height, initial and final ramp spacing
and slip on individual thrusts be assumed (Mitra 1986). By varying these parameters
different geometric forms of thrust systems can be modelled. The progressive
development of a simple hinterland-dipping duplex is illustrated in Fig. 9.10 (Boyer
and Elliot 1982; Mitra 1986; Ramsay and Huber 1987). In the beginning, a major thrust
sheet with flat-ramp-flat geometry and total slip Si develops. The lower and upper flats
define two décollement surfaces, both of which are located at the same stratigraphic
horizon. At the initial stage of duplex formation, a fracture propagates in the footwall
from the lower flat and rejoins the upper flat, tracing a new ramp with initial ramp
spacing as Ri.. Movement along the fracture generates a small horse, which moves
forward by an amount S. The horse climbs up the new ramp and folds the overlying
thrust sheet. The small horse then becomes inactive. Subsequently, a new fracture
propagates from the footwall, the new horse thus formed climbs up the ramp carrying
piggy-back style the earlier formed horse. A roof thrust and a floor thrust form with

84
horses occupying the area in between. The entire process can be repeated as many
times as desired. In this model, the height and angle of the ramps, slip on individual
ramps and spacing between ramps are same during sequential development of ramps.
This leads to rather regular geometry of the duplex. By varying these parameters
different types of duplexes can be generated. For example, by increasing ramp spacing
and keeping all other parameters same we could we could have generated independent
ramp anticlines instead of hinterland-dipping duplex shown in Fig. 9.10. Or, we could
generate an overlapping ramp anticline, antiformal stack or foreland-dipping duplex
by increasing displacements on the thrusts.

Foreland
Major thrust sheet

So Incipient fracture Lower and upper glide horizons

S1 Horse

S2 Roof thrust

S3 Floor thrust

Figure 9.10. Series of diagrams illustrating progressive development of hinterland-dipping


duplexes (after Mitra 1986).

9.6 Thrust sequences

This refers to the sequence of development of thrusts (Fig. 9.11). It is a very


important parameter needed for proper interpretation of geometric and kinematic
evolution of a thrust belt as well as for balancing and restoration of sections.

85
• Breaching: An early-formed thrust cut by a later (i.e., younger) thrust formed in
the hangingwall. The later thrust has the same vergence as the older thrust.
• In-sequence thrusting: A thrust sequence that has formed progressively and in
order in one direction only. There are two types: forward-breaking sequence and
break-back sequence.

(a) 1 2 3

4 3 2 1
(b)

(c) 1 4 2 3

Figure 9.11. Thrust sequences. Numbers indicate the order in which thrusts
developed. (a) Forward-breaking in-sequence thrusts. (b) Break-back in-
sequence thrusts. (c) Out-of-sequence thrusts.

• Out-of-sequence thrusting: Opposite to in-sequence thrusting (Morley 1988). The


out-of-sequence thrusts commonly cut through and displace pre-existing
thrusts.
• Forward breaking sequence: The sequence of thrusting in which new (younger)
thrusts form in the footwalls of older thrusts and all the thrusts verge towards
foreland.
• Break-back sequence: The sequence of thrusting in which new (younger) thrusts
form in the hangingwalls of older thrusts and all the thrusts verge towards
hinterland.

86
• Piggy-back thrust sequence: This sequence occurs when older thrusts are carried
piggy-back style by younger thrusts; essentially same as forward-breaking
thrust sequence.
• Synchronous thrusting: Two or more thrusts move together.

10. Normal Faults


Normal faults are dip-slip faults in which hangingwall moves down relative to
footwall (Fig. 10.1). Absolute movement on a normal fault can rarely be determined
and four possible absolute movements produce the same relative movement:
• hangingwall goes down but footwall remains fixed
• footwall goes up but hangingwall remains fixed
• both hangingwall and footwall blocks move downward but hangingwall
moves more than the footwall
• both hangingwall and footwall blocks move upward but footwall moves more
than the footwall

Well
hangingwall block

fault surface

footwall block
net slip

(a) (b)

Figure 10.1. Normal fault. (a) Hangingwall in a normal fault goes down relative
to footwall. Net slip gives the orientation of the slip vector, which is oriented
down the dip of the fault surface. Younger beds lie above older beds across the
fault surface. (b) In a drill well, normal faults may be recognized from missing
beds.

87
The displacement vectors (i.e., the net slip) in normal faults are oriented almost
down the dip of the fault plane such that they have either zero or negligible strike-slip
component (Fig. 10.1). In previously undeformed sedimentary terrains, younger beds
lie above older beds across fault surfaces where a part of the stratigraphic succession
goes missing. Missing strata in drill wells commonly suggest normal faulting. It should
be remembered that the nature of omission and offset of strata depends on the
orientation of the fault surface relative to the orientation of beds involved in folding.
Therefore, caution must be exercised while interpreting seismic reflection profiles
owing to the fact that a normal fault with zero or negligible dip-slip component may
look similar to a strike-slip component with small dip-slip component. The name
“normal” does not indicate that this type of fault is more common than the other types.
This term originated in British coal mines where such faults were common. If a coal
seam was truncated by a fault, the “normal” practice was to continue the drive for
some distance and sink a shaft to find the missing coal seam. The normal faults are
commonly thought to be steeply inclined with a dip of about 60°. More than three
thousand fault dip measurements from 122 faults in British Coalfields show that 70°
dip (standard deviation = 9°) on normal faults is a better approximation than the
commonly accepted value of 60° (Walsh and Watterson 1988). They are the dominant
structural elements in areas where crustal rocks undergo subhorizontal stretching
accompanied by subvertical shortening, such as oceanic and continental rift zones (Fig.
10.2). The steep dips of normal faults are in conformity with Adersonian model of
faulting. However, it is now widely recognized that low-angle normal faults are also
common in extensional tectonics.
Increased interest in normal faults and extensional tectonics in the last two to three
decades has resulted in a set of complex and often confusing terminology. For a
glossary of normal fault and related terminology see Peacock et al. (2000) and
references therein.

10.1 Horst-and-graben structure

88
Normal faults with about 50°-70° dip commonly occur in conjugate pairs and lead
to the formation of horst-and-graben structure (Fig. 10.2) (Cloos 1955, 1968). The bisector
of the acute angle (about 40-60°) between a pair of conjugate normal faults is sub-
vertical and is parallel to the maximum compressive stress direction. The minimum
compressive stress direction is horizontal and parallel to the bisector of the obtuse
angle between the conjugate faults. Therefore, moderately to steeply-dipping conjugate
normal faults are in conformity with Andersonian model of faulting. Spectacular and
active horst-and-graben structures can be found at mid-oceanic ridges and continental
rift zones such east African Rift. Many inactive and old horst-and-graben structures,
buried under growth or later sediments have been imaged through seismic reflection
profiling (Fig. 10.3).

• Horst: It is an elongated uplifted region bound on either side by sub-parallel


normal faults dipping away from area of uplift. Horsts are commonly bound by
grabens or half grabens (Fig. 10.2a).

Suez
Rift-border fault
Ver tical Intra-rift fault
shortening Syn-post-rift
Pre-rift/basement
Dip domain
Hors t
Graben

Graben N
Sinai
Rift-border
fa ult

Gulf of
Suez
Horizontal
(a) Rift-border
fault
extension
Egypt
Conjugate nor mal
fa ults displa cing
each other

(b)

Figure 10.2. (a) Schematic diagram of horst-and-graben structure in areas of


horizontal extension showing principal structural elements of a rift system. (b)
Map of northern part of Red Sea rift system (after Khalil and McClay 2002).

89
• Graben: It is a long and narrow region of subsidence bound on either side by
sub-parallel normal faults that dip towards each other. Grabens are commonly
bound by horsts (Fig. 10.2a).
• Half-graben: In some terrains, subsidence is only due to one controlling normal
fault and a graben is flanked by only one horst. The structure on the
downthrown side is then termed as half-graben (Fig. 10.4). A half-graben
typically contains a sedimentary wedge that thickens towards the fault.
• Rift zones: These are narrow and elongated zones of subsidence consequent
upon crustal-scale extensions (Fig. 10.2). They are present both on continents
(e.g., East African rift zone) and on ocean floor (e.g., mid-oceanic ridge). Horst-
and-graben structures commonly occur in rift zones. Owing to thinning of the
crust, high heat flow and magmatic activity may be associated with rift zones.

Lr. Tertiary Up. Tertiary


TWT (sec)

2
Palaeozoic
Palaeozoic
3

Figure 10.3. Buried asymmetric graben structure imaged through seismic


reflection profiling, Basin and Range province, Nevada, USA (after Effimol and
Pinezich 1986).

• Rift-border fault: Rift zones are usually bound by a pair of normal faults dipping
towards the middle of the zone. These faults represent the boundary of the rift
zones and are called rift-border fault or basin-margin fault (Fig. 10.2).
• Intra-rift fault: Faults of smaller magnitude than rift-border faults occur inside
the rift zone and are called intra-rift faults (Fig. 10.2).

90
Horst
half-grabens half-grabens

Listric
Detachment fault
fault Detachment
fault

Figure 10.4. Main faults with concave upward listric trajectory gradually
become detachment fault at depth. Subsidiary imbricate faults in the
hangingwall may terminate (left hand side) or merge (right hand side) with the
detachment fault. Half-grabens are bound by only one main fault.

• Compatibility problem: Synchronous movement along a crossing pair of conjugate


system of normal faults leads to the creation of opening between fault walls.
Since rocks are weak materials, large openings can not be sustained in the crust
except within few hundred meters from the surface. This is the so-called
compatibility problem, which can be accounted for if the movement along a set
of conjugate faults is sequential rather than synchronous (Fig. 10.5). Conjugate
normal faults displace each other owing to the compatibility problem (Fig.
10.2a). This is one of the geometric features that can be used to establish if two
sets of faults form a conjugate pair or not.

? ?

1 2
3

Figure 10.5. Compatibility problem associated with steeply-dipping conjugate


normal faults. Synchronous movement along both the faults opens up
unacceptable gaps. Alternate movements along two sets of faults avoid opening
of gaps. Note that fault number 1 is displaced by fault number 2, which in turn
is displaced by fault number 3. Faults 1 and 3 form a single set.

91
• Growth sediments: During normal faulting the downward displacement of
hangingwall block creates low relief relative to the footwall block. Sediments
derived through eroding footwall may fill this low relief and are called growth
sediments (Fig. 10.3).

10.2 Low-angle normal faults

The traditional view of extensional provinces is that their structural characters are
manifestations of high-angle normal faults giving rise to horst-and-graben structure.
This geometry is predicted by Andersonian model of faulting. However, detail surface
mapping and seismic imaging of many extensional terrains show that low-angle
normal faults are also quite common. An initially high-angle normal fault may acquire
gentler dip due to block rotation. Alternately, a steeply-dipping basin boundary fault
may become gradually gentler with depth. As the dip becomes gentler the high-angle
brittle fault near the surface may grade though an intermediate brittle-ductile shear
zone to low-angle ductile shear zones at depth (Ramsay 1980). Such faults with listric
fault trajectory may become or merge with detachment or décollement fault at depth.

10.3 Bookshelf or domino faulting

Strongly rotated fault blocks may be a result of bookshelf or domino model faulting
(Mandl 1984, 1987; Ramsay and Huber 1987). In this type of faulting, a system sub-
parallel faults undergoes progressive rotation of beds and faults as extension continues.
Progressive rotation takes place in such a way that steepening of beds is accompanied
by decrease in dips of faults. Bookshelf faults generally pass into a detachment fault or
into a zone of ductile deformation.

92
Extension fractures

(a)
block rotation

(b)
block rotation
growth sediments new faults old faults

(c)

(d)

Figure 10.6. “Bookshelf” or “domino” model of faulting. See text for discussion.
The initial fractures probably develop as vertical extension fractures (as opposed to
Coulomb fractures) within retaining walls in response to horizontal stretching (Fig. 10.
6a). As the unfaulted walls move away, the faulted blocks collapse sideways and
undergo rotation with normal sense of shearing displacement exactly the same way as
books in a library bookshelf falls sideways if a book is removed (Fig. 10.6b). The angle
of rotation faults is same as the dip of the beds if the rotation of fault block is rigid-
body type and originally beds were horizontal,. The half-grabens formed above the
rotating blocks may be filled up with syntectonic sediments called growth sediments.
The growth sediments are cut by faults and sediments are thicker in the hangingwall
near the fault than in the footwall. This suggests that faulting was active during
sedimentation. Such faults are called growth faults. After a certain amount of rotation, it
may be mechanically more efficient to develop new, steeply oriented fractures rather
than continue to rotate on the old faults (Fig. 10.6c). The bedding planes and old faults
continue to rotate as block rotation is transferred onto new faults (Fig. 10.6d). The old
faults may become sub-horizontal or may eventually dip in the direction opposite to
which they were initiated. They may apparently look like thrust faults but can still be

93
recognized as normal faults because younger rocks in the hangingwall will ride over
older rocks in footwall. This idealized and rather naïve model does reproduce
geometric relations in differently oriented fault blocks commonly found in extensional
basins. However, very large horizontal stretching in extensional basins cannot be
accounted by bookshelf faulting alone because even with high angle of rotation
extension remains rather low.

10.4 Extensional detachment faulting

Low-angle extensional detachment faults (Figs. 10.4, 10.7) are common in many
extensional basins, such as the Basin and Range province of western US (e.g., Davis et
al. 1980). Such faults are not predicted in Andeson’s theory of faulting. The total
extensions in such terrains are typically very high and much of the horizontal
extensions are accommodated along the detachment fault. The hangingwall undergoes
high-angle normal faulting and the variation in dips of strata suggests that the faulted
blocks underwent variable amount of block rotation. The extreme extension often strips
off the unmetamorphosed cover rocks and leads to upwelling of the basement rocks
(Fig. 10.7). Tectonic erosions at such places expose of basement rocks, called
metamorphic core complexes. The detachment faults run close to the contact (but not
necessarily along the contact) between the cover rocks and metamorphic crystalline
basement complex. The basement rocks below the detachment are mylonitized and the
cover rocks above the detachment show cataclastic deformation.

Unmetamorphosed cover rocks

Upwelling of footwall Detachment fault


Mylonitized crystalline rocks due to stripping of
hangingwall

Figure 10.7. Diagrammatic cross-section through metamorphic core complex


showing stripping of fault blocks in the hangingwall exposing underlying
metamorphosed basement rocks.

94
10.5 Listric fault

Normal faults often show smoothly curved trajectory and such faults are called
listric normal faults (Fig. 10.4). Listric geometry allows smooth integration of movement
on a near-surface and steeply dipping normal fault with movement on a detachment
fault at depth. Displacement of hangingwall on listric faults also opens up large gap
and sets up compatibility problem (Fig. 10.8). To solve for the compatibility problem,
different kinds of structures develop in association with listric normal faults (Gibbs
1984):

• Rollover anticline: The displacement of hangingwall on a concave upward listric


fault opens up a large gap, which is not allowed in nature (Figs. 10.8a,b). Beds
in the hangingwall may steepen to form an anticlinal fold and allow the
hangingwall and footwall to remain in contact with each other (Fig. 10.8c). This
fold is called rollover or rollover anticline. If cross sectional area and bed
thickness remains constant during folding then another gap develops in the
hangingwall. Thinning of folded layers and/or shearing in the hangingwall can
close this gap.
• Antithetic fault: Antithetic faults (also called counter faults) are subsidiary to a
dominant fault or fault set and the antithetic faults dip in the direction opposite
to the direction in which the dominant fault dips (Fig. 10.9). The sense of shear
in antithetic faults is opposite to that of the dominant fault. Development of a
set of antithetic faults in the hangingwall may solve to a large extent the
compatibility problem associated with listric faulting (Fig. 10.8d). The antithetic
faults may also have listric trajectory.

95
gap

(a) (b)
Fold Faults

(c) (d)

Figure 10.8. Compatibilty problem in listric fault (after Twiss and Moores 1992).
(a) Incipient fracturing of with listric trajectory. (b) Movement of hangingwall
opens up a large and geologically untenable gap. (d) Compatibility problem
may be solved through distribution deformation in the hangingwall leading to
the formation of rollover anticline. (d) Compatibility problem may also be
solved through antithetic faulting in the hangingwall. Note that small gaps are
present below the faulted blocks.

Graben
Main
fault Main
fault

Synthetic
faults Antithetic
faults

Figure 10.9. Antithetic and synthetic faults associated listric main faults in a
graben.

• Synthetic faults: These faults are minor or subsidiary faults with same sense of
shear and similar orientations as the related major fault (Fig. 10.9). Antithetic
and synthetic faults may form a conjugate pair.
• Rider: Formation of a second, and then sequential subsidiary faults cutting back
into the undeformed footwall give rise to wedge-shaped segments in cross-
section between faults. These are called riders (Fig. 10.10a). The riders develop
in-sequence towards the footwall (Fig. 10.11). Each rider is bound by a roof fault
and a floor or sole fault. The roof faults on riders passively rotate on the lower
active fault. Riders can be both synthetic and antithetic. At high extensions,
riders can get detached from each other may look like horst-and-graben
structure in seismic sections (Figs. 10.10c,d).

96
riders only on footwall riders only on hangingwall
8
Ri Hangingwall 1 7
Footwall de 2 6
Sol r R i R oo 3 4 5
eo d e f fa
rF r u lt
loo (b)
(a) r fa
ul t

Riders Growth sediments


Detached riders at
high extensions

(c) Sole fault


(d)

Figure 10.10. Structures associated with listric faults.

• Listric fan: A set of synthetic riders form a listric fan, also called horsetail faults
(Fig. 10.10a).
• Reverse listric fan: A set of riders antithetic to the main listric fault but rest on the
hangingwall is called reverse listric fan (Fig. 10.10b).
• Roof fault: The topmost fault in a listric or reverse listric fan (Fig. 10.10a).
• Floor fault: The lowermost fault in a listric or reverse listric fan (Fig. 10.10a).
• Extensional duplex: An extensional duplex is made up of stacked up horses and
bound by a roof fault and floor fault (Fig. 10.11). An extensional duplex formed
in association with listric normal faulting has geometry similar to a duplex
formed during thrusting. Only sole fault in an extensional duplex is active and
the roof fault consists of sequential faults. They commonly form where the main
fault has ramp-flat trajectory.

97
Central ridge
5 Listric fan
4 Counter
3 fan 4
3
2
1 1 2

4
3 2 1
Extensional duplex

Figure 10.11. A composite diagram showing listric fan, counter fan and
extensional duplex. The numbers indicate sequence of fault development.

98
99
11. Strike-Slip Faults
Strike-slip faults are commonly vertical faults with displacement predominantly in
the horizontal direction (Fig. 11.1a). The dip-slip component is zero or small compared
to strike-slip component. The sense of displacement of a strike-slip fault is either
clockwise or counterclockwise and they are called dextral (syn. right-handed or right-
lateral) or sinistral (syn. left-handed or left-lateral), respectively. If we stand on a fault
and face one of the blocks, then if the block facing us has moved to right then it is right-
handed (dextral) and it is left-handed (sinistral) if it has moved to the left (Fig 11.1b).
Strike-slip faults are also known by several other names, such as lateral-, wrench-, tear-
and transcurrent fault. The descriptive meaning of the term "strike-slip" is definitive
and simple and, therefore, should be preferred over all other names.

sinistral
net slip

fault surface

(a) (b) dextral

Figure 11.1. (a) Strike slip fault showing displacement vector, which is
essentially horizontal. (b) Sinistral (left-handed) and dextral (right-handed)
sense of displacement.

11.1 Transform fault

Transform fault is a special type of strike-slip fault, most spectacularly seen on an


ocean-floor map cutting across mid-oceanic ridges. These faults link (or transform)
other major tectonic features along plate boundaries, such as graben systems of mid-
100
oceanic ridges (Fig. 11.2). There are three unusual aspects about displacement on
transform faults: (1) displacement remains constant throughout the fault and then ends
abruptly, (2) the actual sense of displacement is opposed to what can be inferred from
offsets of geologic features, and (3) two sub-parallel faults may have opposite sense of
movements. The offset graben system in Fig. 11.2 suggests sinistral movement but the
actual displacement is dextral.

transform fault

sense of diplacement
opposite to ridge offset

mid-oceanic
rift system

Figure 11.2. A strike-slip fault joining (or transforming) mid-oceanic rift


systems. Such strike slip faults are called transform faults. Note that the actual
sense of displacement is right-handed although offset rift suggests left-handed
displacement. The right-handed displacement on the transform fault is in
conformity with the extension at the rift.

11.2 Transpression and transtension zones

A strike-slip fault or a set of parallel strike-slip faults affecting a set of sub-


horizontal beds do not generate much subsidiary structures. Consequently, such faults
may appear to be rather uninteresting. However, strike-slip faults may occur in en-
echelon sets or they may have bends (or jogs). Displacement in such situations results in
the formation of strike-slip duplex, which is a set of horizontally stacked horses bound
on both sides by parts of the main strike-slip faults. They are also called flower-, tulip- or
palm-tree structure. In an en-echelon set, a set of parallel strike-slip faults are oriented at
the same angle to a reference line, called bearing line (Fig. 11.3). Adjacent faults in en-
echelon sets show right (Figs. 11.3a,b) or left (Figs. 11.3c,d) sense of strike shift.
Similarly, curved strike-slip faults may be considered to have right-handed (Figs.
101
11.4a,b) and left-handed (Figs. 11.4c,d) jogs. These right- and left-handed senses of shift
and sense of shear on fault surfaces lead to two situations:

line of bearing

(a) (b)

(c) (d)

Figure 11.3. En-echelon strike-slip fault system. All the faults in an en-
echelon system are inclined at a constant angle to a reference line, called
line of bearing. The sense of shift of adjacent faults is either right-handed
(a, b) or left-handed (c, d). The sense of shear displacement on the faults
are either left-handed (a, c) or right-handed (b, d).

(a) (b)

(c) (d)

Figure 11.4. The strike-slip faults with bends (or jogs) in strike line. (a,b)
Right-handed jog. (c,d) Left-handed jog. (a, c) Right-handed shear
displacement. (b, d) Left-handed shear displacement.

• The shear sense on fault and sense of en-echelon shift or jog are same (Figs.
11.5a,b). In this case, translation due to faulting is accompanied by extension set
up in the zone of overlap between en-echelon faults or in the zone where fault
trace is curved. Such combination of translation and extension is termed as
transtension. Structures typical of extensional deformation may develop in
transtension zones. The most common large-scale structure is pull-apart basin,
which is a rhomb-shaped graben (Fig. 11.6.. Normal faults in a pull-apart basin
are oriented initially at about 45-50° to the strike-slip fault. This angle increases
102
as normal fault block rotates with increasing shear. These basins may get
sediments derived through erosions and denudation of topographically higher
regions bounding the graben. On a smaller scale, local depressions may lead to
the formation of sag ponds, which may be site of temporary or permanent lake.
The normal faults in the graben may have concave upward shape and merge
with the main strike slip fault. The faults may have significant amount of both
strike-slip and dip-slip components. Such a structure is called normal (or
negative) flower (or tulip) structure (Fig. 11.7a).

zone of
transtension

(a) (b)

zone of
transpression

(c) (d)

Figure 11.5. (a,b) Zones of transtension (translation + extension) developed due


to same sense of strike-slip displacement and sense of en-echelon shift or jog.
(c,d) Zones of transpression (translation + compression) developed due to
opposite sense of strike-slip displacement and sense of en-echelon shift or jog.

normal sediment
faults deposit

(a)

thrust fold axial


fault trace

(b)

Figure 11.6. Structures develop in zones of transtension (a) and transpression


(b).

103
• The shear sense on fault and sense of en-echelon shift or jog are opposed (Figs.
11.5c,d). A combination of translation and compression follows and a zone of
transpression forms. The compression may to local vertical uplift of a rhomb-
shaped region, called pressure ridges. On a larger-scale, the compression may
lead to the formation of folds and thrusts. The axial traces and fault traces will
be oriented initially at 45-50° to the strike-slip faults. With increasing shear
along the main strike-slip fault, the axial and fault traces may rotate towards the
main fault. The thrust faults with concave downward shape may also have
significant strike-slip component. These thrust faults may have the geometry of
a reverse (or positive) flower structure (Fig. 11.7b).

(a) (b)

Figure 11.7. (a) Normal or negative flower structure. (b) Reverse or positive
flower structure.

11.3 Slip rate

The rate at which fault blocks move past each other may not be uniform and frontal
part of fault may move at a different velocity than the rear portion (Fig. 11.8). This
leads to a compatibility problem, which are solved through the development of
subsidiary structures. If the frontal part moves at a faster rate than the rear part then
the fault block will be stretched and extensional structures such as normal fault may
form. If, on the other hand, the frontal part moves at a slower rate than the rear part

104
then the fault block will be compressed and structures such as reverse faults or folds
may form.

t er
r
r

ste
we

F as
Fa
o
Sl

r
we

rwe
Slo

Slo
r
ste
Fa

Figure 11.8. Diagram showing zones of extension and compression developed


due to different slip rates at the front and rear of a fault block.

R2

Main strike-
slip fault

P R1

Figure 11.9. Subsidiary shear fractures developed in association with right-


handed strike-slip fault.

11.4 Subsidiary shear fractures

The strike-slip faults may be associated with wide range of subsidiary shear
fractures. The most important of these are Riedel shears or R shears. They were originally
recognized in laboratory experiments in which a layer of clay was deformed overlying
a sharply defined vertical strike-slip fault generated by two rigid blocks sliding past
each other. Two sets of en-echelon shear fractures developed in the clay layer, one
lying at 10-15° and the other at 75-80° to the underling fault surface (Fig. 11.8). They are
designated as R1 and R2, respectively. The R1 has the same sense of movement as the
105
main strike-slip fault, but the R2 has opposite sense of movement. That is, if the main
strike-slip fault has dextral sense of shear then the R1 and R2 shears have dextral and
sinistral sense of displacement. In other words, the R1 and R2 are synthetic and
antithetic respectively. Another type of subsidiary shear, called P shears, synthetic to
main fault and symmetrical to R shears may also develop. Subsidiary shears on large
scale can form a complex and anastomosing network of faults. The geometry and
kinematics of such a complex network of faults may be difficult to interpret.

106
107
12. Folds
Folds are regular wavelike undulations traced by sideways deflections of layers or
surfaces in rocks. They occur in scales ranging from microscopic, through outcrop or
mountain sides and cliffs to tens of km in orogenic core zones. Folds are very common
in metamorphic tectonites formed in response to ductile deformation in the deeper part
of the crust. They also form in the shallow crustal depths in sedimentary or very low-
grade metamorphic rocks in the flanks of major orogenic zones. The terminology for
the purpose of geometric description of folds has evolved over a long period of time.
The terminology is rather extensive, not always consistent and some of them have
genetic implications. However, we recognize that descriptive terminology should be
devoid of any genetic implications. Folds are traced by layers, such as sedimentary
beds, veins, dikes and metamorphic and igneous bands. Folds may also be considered
to have been traced by single surfaces, such as bedding and cleavage surfaces. The
geometry of single folded surfaces should be treated separately from the geometry of
folded layers.

12.1 Curvature of a folded surface

A curved line or surface can always be represented by a mathematical equation of


the form z = f(x) or z= f(x, y). The first differential of the equation gives the tangent (line
or plane), which is a measure of slope at a point. The second derivative gives the
variation in the orientation of the tangent and represents curvature at a point. For
example, a straight line has zero curvature and a circular arc has constant curvature.
The following terms are used to describe the variation in curvature of a single folded
surface (Fig. 12.1):

108
• Hinge line: The hinge line is the line along which the folded surface has
maximum (positive or negative) curvature. On a cross section, hinge line will
obviously be represented by a point, called hinge point. Both hinge line and
hinge point are simply called hinge. The hinge line can also be thought of as loci
of points of maximum curvature. The hinge line of a fold need not be a straight
line; it may be horizontal or plunging.

fold limb
hinge fold domain
zone
e
li n
e
ng
e

crest point hi
li n
n
iox

i ne
fl e

hinge
in

el
point
ng
hi

inflexion
point fold domain
trough point hinge point

hinge line
depression

hinge line
culmination

kink fold e concentric fold


e li n hinge point
hing
hinge point

hinge point
e
e li n
h in g
hinge point

Figure 12.1. Geometry of single folded surface.

• Inflexion line: Inflexion lines mark zero curvature on a folded surface. The sign
of curvature changes (i.e., the sign of the second derivative) across the inflexion
lines. In other words, the sense of curvature of the folded surface changes, for
example from convex up to convex down. On a section, inflexion line is
represented by a point, called inflexion point.

109
• Crest and trough: Crest and trough (line or point) of a fold represent highest and
lowest, respectively, topographic elevations on a folded surface. Crest and
trough may, but not necessarily, coincide with hinge.
• Hinge line culmination and depression: If the plunge of hinge line varies, then there
should be areas where hinge line will attain highest and lowest elevations,
called hinge line culmination and depression, respectively.
• Crest/trough line culmination and depression: Crest lines or trough line may also
have culminations and depressions, as with hinge lines.
• Fold axis: It is an imaginary line on the folded surface, which when moved
parallel to itself generates the folded surface. Such folds are also called
cylindrical folds. The non-cylindrical folds surfaces do not have this property.
Hinge lines are straight in cylindrical folds. The terms fold axis and hinge line
are sometimes used synonymously, although they are not the same.

12.2 Fold domain

The folded surface between two successive inflexion lines is called a fold domain
and constitutes one fold (Fig. 12.1). Hinge zone and limbs are parts of a fold domain.
• Hinge zone: The surface region around the hinge line where the curvature is
high is called a hinge zone. Although a single hinge per fold domain is common
in most folds, a single fold domain can have more than one hinge.
• Fold limb: The surface region between inflexion line and hinge zone is known as
limb of a fold. Fold limbs can be straight or curved. A fold domain commonly
has two limbs.
• Concentric fold: If the fold domain is part of a perfect circular arc, then the fold is
called a concentric fold. Hinge points in such cases are taken at the midpoints of
each of the fold arcs.
• Kink fold: Folds with straight limbs and sharp hinge are called kink or chevron
folds. The curvature at the hinge is infinite and at limbs it is zero. Inflexion
points are taken at midpoints on the straight line segments of the folds.

110
Neutral
antifom

synform
Vertical

Younger Older
synformal synformal
syncline anticline
Older Younger
Younger
antiformal antiformal Older
anticline syncline
Older Younger
(e)

Monocline inverted saddle sheath fold


saddle
hinge line

hinge line hinge line


(d)
(k) non-cylindrical fol ds

Figure 12.2. Geometric terms based on closure of fold domains.

12.3 Fold closure

Following is a set of names given to folds depending on the closure of fold domains
(Fig. 12.3):

• Antiform: If the fold domain closes upward (-ve curvature), then the fold is
called antiform.
• Anticline: In an anticline older rocks are located in the core of the fold.
• Synform: In a synformal fold, the fold domain closes downward (+ve closure).
• Syncline: If younger rocks are located in the core of the fold, the fold is syncline.
• Monocline: Regional step-like folds in which otherwise horizontal or shallowly
dipping strata abruptly bend to steeper inclination within a very narrow zone.
• Homocline: A set of uniformly and gently dipping beds constitute a homocline.
• Neutral fold: When the fold domain closes sideways, the folds are called neutral
fold.
111
• Vertical fold: A type of neutral folds in which both hinge lines and limbs are
vertical.
• Overturned fold: If the two limbs of a fold dip in the same direction, it is called
overturned fold. In this type of fold, one of the limbs rotates more than 90° from
initial horizontal orientation.
• Dome: A dome is an antiformal fold domain with hinge line culmination. Such a
fold has an approximate shape of a dome.
• Basin: A dome is a synformal fold domain with hinge line depression. Such a
fold has an approximate shape of a basin.
• Saddle: A type of antiform with hinge line depression giving rise to shape of a
saddle.
• Inverted saddle: A type of synform with hinge line culmination.
• Sheath fold: If the hinge line curves more than 90°, the fold domain may acquire
the shape of sheath of a knife or sword. A fold with this shape is called sheath
fold.

Combination of antiform/anticline and synform/syncline gives four possible


geometry, viz., antiformal anticline, synformal syncline, antiformal syncline and
synformal anticline. The geometries of antiformal syncline and synformal anticline
require inverted stratigraphy. In deformed sedimentary terrains stratigraphy is rarely
inverted and antiform/anticline and synform/syncline are, and can often be, used
interchangeably.

12.4 Fold tightness

The angle between tangents to the folded surface drawn through successive
inflexion lines is called interlimb angle (Fig. 12.3). The interlimb angle provides the
degree of tightness of a fold. Following names are given to folds on the basis of
interlimb angle (Fleuty 1964):

• Gentle: Interlimb angle 180° - 120°


• Open: Interlimb angle 120° - 70°
112
• Close: Interlimb angle 70° - 30°
• Tight: Interlimb angle 30° - 0°
• Isoclinal: Interlimb angle 0°
• Elastica: Interlimb angle negative

The above terms cannot be used in combination. For example, a fold cannot be tight
isoclinal but the tightness of a set of folds can range from tight to isoclinal.

inter-limb
angle 180o
120o
tangent gentle 70o
tangent
at i1 at i2
open
close 30o
inflexion tight
inflexion i2 point 0o
point i1 isoclinal

Figure 12.3. Interlimb angle and tightness of fold (after Fleuty 1964).

12.5 Folded layers

Geometry of a folded multilayer is equivalent to the geometry of a set of stacked up


folded surfaces. Several terms are used to describe the geometry of folded multilayers,
the most import being the axial surface (Fig. 12.4):

• Axial surface: Surface formed by joining successive hinge lines in a multilayered


sequence is called axial surface. Axial surface need not be planar in a
multilayered sequence but when this surface is planar, it is called axial plane. It
is important to remember that axial surface is not defined for a single folded
surface and axial surface need not necessarily divide a fold into two symmetric
halves.
• Axial trace: Trace of the axial surface on the topographic surface. The term axial
trace has nothing to do with fold axis. Axial trend (i.e., the trend of hinge line)
and axial trace may or may not be parallel.

113
• Inflexion surface: Inflexion surfaces are formed by joining successive inflexion
lines in a multilayered sequence. A fold domain is bound by two adjacent
inflexion surfaces. When two inflexion surfaces bounding a fold join, the fold
ceases to exist.
• Conjugate fold: A fold domain may have converging pair of axial surfaces and
such folds are called conjugate folds. Obviously, conjugate folds have two
hinges. The conjugate fold changes into a single-hinged fold where the two
axial surfaces meet.
• Polyclinal fold: In this type of fold, there are more than two hinges in one fold
domain. Such folds are however rather rare.

inflexion line surfaces


axial surfaces
hinge line

two axial surfaces


merge to give one
axial surface

hinge line
axial plane fold ceases to exist
at this point

Figure 12.4. Axial surfaces of folds.

12.6 Fold symmetry

The following terms are associated with symmetry of folds (Fig. 12. 5):
• Fold train: A fold train is a series of folds with alternating sense of curvature.
• Wavelength: Wavelength of a continuous and regular waveform is defined as the
distance between two successive maxima or between any two points in the
same phase. However, a fold train rarely has the geometry of a regular
waveform and wavelength cannot be measured directly. The distance between
two successive inflexion points in a fold domain can be measured and this
distance is the half wavelength of the fold.

114
• Amplitude: Amplitude is the perpendicular distance between the median line
and the point on the folded surface that has the maximum deflection from the
median surface.

Fold
domain 1
A
W/2
W/2 median
A surface

axial plane
Fold asymmetric fold
A: Amplitude domain 2 symmetric fold
W: Wavelength
parasitic fold
harmonic folds disharmonic folds

median surface

anticlinoria-synclinoria

Figure 12.5. Fold symmetry.

• Median surface: It is a surface formed by joining successive inflexion lines.


• Symmetry: If the median surface and the axial surface are perpendicular to each
other and the axial surface divides the fold domain into mirror symmetric
quarter halves, the fold is symmetric. Otherwise the fold is asymmetric. The
symmetric folds are m-shaped and asymmetric folds are either s-shaped or z-
shaped.
• Harmonic/disharmonic folds: If the wavelength, amplitude and geometry of folds
in a multilayered sequence are similar along the axial surface then the folds are
harmonic, otherwise folds are disharmonic. The variation in the geometry of
folds across axial surfaces has nothing to do with this geometric description.
• Polyharmonic fold: If a fold train has two or more orders of folds it is called
polyharmonic fold. Each order of fold has its own characteristic wavelength and
amplitude.

115
• Parasitic folds: In a polyharmonic fold, the folds with smallest wavelength and
amplitude are called parasitic folds.
• Anticlinorium: It is a very large anticlinal polyharmonic fold.
• Synclinorium: It is a very large synclinal polyharmonic fold.

12.7 Fold attitude

The orientation or attitude of folds in three-dimensional space is important for


proper description of folds. Fleuty (1964) proposed a complete set of descriptive name
of folds based on the dip of axial plane and plunge of hinge line (Figs. 12.6, 12.7):

• Upright: Folds with sub-vertical dip (>90°) of axial planes are called upright.
Depending on the plunge of the hinge line, the upright folds are pre-fixed with
sub-horizontal, gently-plunging, moderately-plunging and steeply-plunging.

Dip of axial surface

Sub-horizontal
Moderately
inclined

inclined

inclined
Upright

Steeply

Gently

0 0 0 0 0 0
90 80 60 30 10 0
00
Sub-horizontal
0
10
Recumbent
Gently
plunging

ed
Plunge of fold hinge

300 in
n cl
I
Upright

Moderately
plunging

600

Steeply
plunging
Reclined
800
Sub-vertical
90
0 Vertical

Figure 12.6. Names of folds based on the orientation of axial plane and hinge
line (after Fleuty 1964).

116
• Vertical: Folds with sub-vertical (>90°) dip of axial plane and plunge of hinge
line are called vertical folds. As a consequence of this geometry, the folded
surface is also sub-vertical everywhere.
• Recumbent: Folds with less than 10° dip of axial plane as well as plunge of hinge
line are called recumbent folds. The limbs of recumbent folds are also sub-
horizontal but dip at the hinge zone is vertical.
• Reclined: In a reclined fold, dip direction of the axial surface is the same as the
trend of the hinge line. An equivalent statement is that the hinge plunges down
the dip of the axial surface.
• Inclined: In this type of fold the dip of axial plane and plunge of hinge line vary
between 10-80°.

(a) Upright (b) Vertical

(c) Recumbent (d) Inclined horizontal

(e) Upright moderately inclined (f) Moderately inclined moderately plunging

(g) Reclined

Figure 12.7. Three-dimensional geometry and stereographic projection


diagrams of different types of folds based on the attitude of folds as defined by
Fleuty (1964).

117
12.8 Fold classification

Folds can be very precisely classified on the basis of variation of thickness.


Historically, the description of variation in layer thickness developed around two
geometric models, viz., parallel fold and similar fold (Fig. 12.8) (van Hise 1896). In parallel
fold the orthogonal thickness (i.e., thickness measured orthogonally across the layer)
remains constant throughout the fold. These folds are usually found in competent
layers surrounded by less competent matrix. Parallel folds do not continue for long
distances along the axial surfaces but die along a décollement surface. In similar folds
the thickness measured parallel to the axial plane remains constant throughout the
fold. The orthogonal thickness in similar folds vary considerably with the limbs
thinned and the hinges thickened. A special property of similar folds is that they can
continue indefinitely along axial surfaces. Most of the natural folds show significant
departure from these two ideal end member models.

Class 1A
t t0 = T0
T Class 1B Parallel
1.0

Class 1C
t'
t - Orthogonal thickness at dip
Cl
ass

T - Axial Planar thickness at dip


2

t / t0
0.5
Sim

t' =
i la

Class 3
r

0.0
0 0 0
0 30 60 90

Figure 12. 9. t'-α classification of folds (after Ramsay 1967).

118
Ramsay (1967) gave a very precise scheme of classification of folds based on
variation in layer thickness in quarter wavelength of folds (Fig. 12.9). Two parameters
are defined, viz., limb dip (α) with respect to a datum taken as tangent at the hinge
point and thickness (tα') as a ratio between thickness measured at the hinge and at α
limb dip. Three classes are recognized: class 1 with tα' > cos α, class 2 (similar fold) with
tα' = cos α and class 3 with tα' < cos α. Class 1 folds are further divided into class 1A
with tα' >1, class 1B (parallel fold) with tα' = 1.0 and class 1C with tα' <1 (but > cos α).
This classification is also compatible with dip isogon classification of folds (Fig. 12.10).
Dip isogons are lines joining points of equal dip (measured with respect to the datum)
measured on either sides of the fold. Fold classes 1, 2 and 3 have convergent, parallel
and divergent dip isogons, respectively. In classes 1A and 1C, the isogons are strongly
and weakly convergent, respectively. Dip isogons in class 1B folds are oriented
perpendicular to layering. All measurements and constructions are carried out on
profile section, i.e., sections perpendicular to hinge lines.

Class1, convergent isogons, inner curvature > outer curvature

1A 1B, Parallel 1C
strongly convergent isogons perpendicular weakly convergent
isogons to layering isogons
Class 2, Similar Class 3

-dip isogon

parallel isogons, diverging isogons, -dip isogon


curvature of two curvature of outer arc
surfaces same > inner arc

Figure 12.10. Dip-isogon classification of folds (after Ramsay 1967).

119
12.9 Kink folds

Kink fold geometry is commonly used for the purpose of structural modelling in
sedimentary terrains, such as fold-thrust belts although perfect kink geometry is rare in
nature. However, many folds in these terrains have straight limbs and sharply curved
hinges of small areal extent. No significant error is introduced while modelling if such
folds are treated as kink folds. Advantages of kink fold geometry include dip panels
can be easily constructed, axial surfaces can be unambiguously defined and easy
mathematical description of fold.
In kink and chevron folds (Fig. 12.11), the limbs are straight and hinge zones are
sharp (i.e., angular) with infinite curvature (because radius of curvature is zero).
However, the hinge zones in natural kink/chevron folds may not be perfectly angular
but may be slightly curved within a small areal extent. This type of fold occurs in
strongly anisotropic rocks such as thinly-stratified sedimentary rocks or low- to
medium-grade metamorphic rocks with slaty cleavage/schistosity.
The terms kink fold and chevron fold are not synonymous but they are closely
related. Chevron folds are roughly symmetrical with approximately equal limb lengths
and have one set of sub-parallel axial surfaces. fold train is usually continuous across
the axial surfaces for some distance. kink folds, on the other hand, are asymmetric with
markedly different limb lengths. There is usually a discrete zone in the rock within
which the layers have different orientation as compared to the orientation outside. In
an ideal kink fold (Figs. 12.11; Fail 1969), dips of layers abruptly change across an
imaginary planar boundary, called a kink plane, which represents axial plane. The kink
axis (or hinge line) is given by the intersection between kink plane and folded surface.
A kink band is the area between two adjacent kink planes whose kink axes are parallel.
The line along which two kink planes join is called kink junction axis. If the kink junction
axis lies in the plane of layering then the fold is cylindrical, otherwise the fold is non-
cylindrical. If layers do not change thickness, the kink plane bisects interlimb angle (i.e.,
γ1 = γ2, in Fig. 12.11).

120
kink band
kink plane
Kink axis

axial
planes

Kink axis

1 1
2 2
T1 t1 T1 t1
t2 axial
T2 t2
planes α α
T2

Figure 12.11. Geometric characteristics of kink folds.

Kink folds are peculiar in the sense that the folds are both similar and parallel. If
the two limbs have different thickness, then γ1 ≠ γ2 but individual limbs still have the
geometry of both parallel and similar folds. There is a region near the hinge where
orthogonal thickness cannot be defined.
There is only one set of sub-parallel kink planes in monoclinal kink bands. However,
kink bands may occur in two differently oriented sets producing conjugate kink bands. In
other words, two sets of kink bands may be inclined towards each other. When two
kink planes meet they annihilate each other and only one kink plane extends from the
kink junction axis (Fig. 12.11). This property can be used to construct cross section in
areas where kinking is the preferred mode of deformation.

121
13. Modelling of Fault-related Folds
Rigid-body translation of faulted blocks along non-planar fault surfaces leads to
creation of voids (Figs. 13.1a, b). As noted earlier, large voids cannot be sustained in the
crust. In order for the two faulted blocks across a non-planar fault surface to remain in
tight contact with each other, there must be distortions within at least one of the fault
blocks. Layered rocks may develop folds when they slip past a bend on a faulted
surface (Figs. 13.1c, d). The fold geometry is controlled by the shape of the fault. These
folds are called fault-related folds because folding is a consequence of faulting. Such
folds are common in deformed sedimentary basins in both compressional and
extensional tectonic settings. The emphasis here is on thrust-related folds because they
have been described more frequently from fold-thrust belts.

gap
(a) Fracture (b)

(c) gap (d)

Figure 13.1. Compatibility problem arising out of a rigid-body


translation of hangingwall block along a fault with a sharp bend. (a)
Incipient fracture with a sharp bend. (b) Gaps open up due to rigid-body
translation of hangingwall block. (c, d) A kink band develops and fills
up the gaps. The geometry of the kink band is controlled by the shape of
the fault. Note that the gap above the footwall on the left side of (d) is
yet to be filled up.

Since we do not understand mechanics in sufficient detail as yet, fault-related folds


are usually modelled kinematically and the geometric consequences are evaluated

122
against naturally occurring folds and related structures. "Kinematics" is a branch of
mechanics that treats motion in an abstract framework, without reference to force and
mass. Since the publication of a classic paper by Suppe (1983), a number of
kinematically valid fault-related folding models have become available. However, fault-
bend folding, fault-propagation folding and detachment folding may be considered as end-
member models and they will be described in detail in this section.

13.1. Assumptions

For the purpose of kinematic modelling of fault-related folds several reasonable


assumptions (or boundary conditions) must be made. This is a legitimate exercise for
any kind of modelling.

• Fault shape: Faults are taken to have sharp bends leading to “ideal” stair-case
trajectory although fault bends may not be as sharp as bends in ideal stair-case
trajectories. Flat-ramp trajectories of thrust faults have approximately stair-case
geometry. The fault bend controls the location and initiation of axial surfaces of
fault-related folds.
• Fold shape: Folds are assumed to have kink fold geometry. Straight limbs and
sharp hinges with infinite curvature of kink folds make the model calculations
easier and simple. This is not a bad assumption because the folds in deformed
sedimentary terrains usually have straighter limbs and hinges of small areal
extent that approximate kink-fold geometry. The assumption of kink fold
geometry does not introduce large error so long as axial surfaces and dip panels
can be unambiguously determined.
• Layer thickness: Layer thickness, bed length and cross sectional area are assumed
to remain constant during folding. In some models thickness of limbs are
allowed to change during folding.
• Deformation: Plane strain is assumed, i.e., the material points are not allowed to
move in and out of section plane. Deformation is essentially accomplished by
slip parallel to bedding with or without simple shear perpendicular to bedding.

123
Rocks do not undergo any internal deformation, i.e., deformation is constant
volume.

13.2 Kinematics of kink folds

Kink folding with constant layer thickness is commonly accomplished by flexural-


slip mechanism wherein layers slip past each other (Figs. 13.2, 13.3). The type of layer-
parallel slip in kink folds developed through layer-parallel compression (i.e., buckling;
Fig. 13.2) is different from those developed due to faulting (Fig. 13.3). In buckling, the
sense of slip on two limbs are opposite in the two limbs and amount of slip decreases
from a high value near inflexion surface to zero at the axial surface Fig. (13.2). Material
points cannot move past the axial surface.

dextrally sinistrally
sheared vein sheared vein

Axial
surface

(a) (b)

Figure 13.2. Flexural-slip due to layer-parallel compression. Sense of shear on


the two limbs is opposite to each other. The amount of slip (i.e. shear strain)
decreases to zero at the hinge. Consequently, the material points cannot pass
through the axial surfaces.

Any pre-existing vein will be displaced on the limbs and shear strain will be
positive and negative in adjacent limbs (Fig. 13.2b). Further, sense of shear can be used
to locate the hinge of the fold. The location and orientation of axial surfaces in fault-
related folds are controlled by bends on the fault surface (Fig. 13.3). Sense of shear
displacement will be same on both the limbs and material points can move past axial
surfaces. Any pre-existing vein will show same sense shear displacement everywhere
(Fig. 13.3c). Ramp portions will always have shear but the flat portions may or may not

124
have shear strain. In either case, if the layers across the axial surface maintain constant
layer thickness, the axial plane must bisect the interlimb angle.

post-deformation
position pre-deformation
position
material point

ψ
ψ ψ' ψ

(a) (b)
sinistrally
sheared vein

(c)

Figure 13.3. Flexural-slip during fault-related folding. The material


points can roll through axial surfaces (a, b). (a) Shear strain is set up on
the ramp part although there is shear on the flat part. (b) Shear strain in
the ramp part is different from shear strain on the flat part. (c) Sense of
shear displacements on the two limbs of the kink fold is same, as given
by sheared veins.

13.3 Fault-bend folding

In the model of fault-bend folding (Suppe 1983), a fracture with a staircase or flat-
ramp-flat trajectory forms rapidly followed by movement of one or both the fault
blocks. If the rocks are layered they may fold in response to riding over a bend in the
fault; the folds thus formed are called fault-bend folds. If we move the hangingwall
keeping the footwall fixed, a flat-crested anticline forms over the ramp and the fold is
called a ramp anticline (Figure 13.4). Names of different parts of ramp anticline and the
angular parameters used for kinematic modelling of fault-bend folding are shown in
Fig. 13.4.
Progressive development of a fault-bend fold (FBF) caused by a simple step in
décollement with folding confined to the hangingwall block is shown in Fig. 13.5
(Suppe 1983). At the time of initiation of folding two axial planes B and B’ form at the

125
lower bend Y and two axial planes A and A’ form at the upper bend X. With continued
slip, the axial plane B’ climbs up the ramp and axial plane A’ moves along the upper
flat. The material points roll through axial planes A and B. The axial planes A and B
do not move. Since the footwall is fixed and only hangingwall moves, we can say that
axial planes A and B are attached to the footwall and axial planes A’ and B’ are
attached to the hangingwall. As the axial planes B’ climbs up the ramp, the fold
amplitude increases but the width of the flat crest is reduced. The width of the two
kink bands AA’ and BB’ also increases with increasing slip. When Y’ reaches X (upper
bend), the axial plane B’ gets attached to the footwall and stops moving. At the same
instant, the axial plane A is transferred to the hangingwall and starts moving along the
upper flat. The ramp anticline stops growing in amplitude but width of the flat crest
keeps increasing with continued slip.

Transport
trailing
direction leading
anticline anticline
trailing backlimb ramp forelimb leading
syncline anticline syncline
δ
trailing
γ leading
γ
edge edge

φ β

Figure 13.4. Different parts of and angular parameters in a ramp


anticline. θ - initial cut-off angle, β - final cut-off angle, φ - change in the
dip of fault (fault shape), γ - half of interlimb angle (Fold shape), δ -
change in dip across axial surface.

Suppe (1983) recognized several angular parameters, which can be used to describe
the fault and fold geometry (Fig. 13.4): change in dip of fault (φ), axial angle (i.e. half-
interlimb angle) of fold (γ), initial cut-off angle (θ), final cut-off angle (β), and change in
dip across axial surface (δ = 180° - 2γ). If the lower flat is parallel to bedding then θ is
also the step-up angle. The angular parameters φ and γ represent fault and fold shapes
respectively. For a simple step from one décollement to another (i.e., θ = φ), γ is related
to θ by the following equation:

126
⎡ sin 2 γ ⎤
φ = θ = tan −1 ⎢ 2 ⎥
(13.1)
⎣1 + 2 cos γ ⎦

For a general case (θ ≠ φ), θ, φ and γ are related by the following equations:

⎡ − sin( γ − θ) [sin(2 γ − θ) − sinθ] ⎤


φ= ⎢ ⎥ (13.2)
⎣ cos( γ − θ)[sin(2 γ − θ) − sin θ] - sinγ ⎦
β = θ - φ + (180° - 2γ) = θ - φ + δ , where δ = 180° - 2γ (13.3)

A B' B
A'

X
X'

a Y' Y

A B'

A' B

X' X
Y'

b Y

A B'

A' B

X’ Y' X

c Y

Figure 13.5. Progressive development of fault-bend fold (after Suppe


1983)

Suppe (1983) provides a graph (Fig. 13.6), which is a pictorial representation of eqs.
13.1–13.3. The graph allows a quick analysis of possible range of solutions to a given

127
problem. For example, for θ = φ maximum cut-off angle cannot be greater than about
30°. This suggests that initial dip of most thrust faults should be 30° or less. Another
interesting point to note is that for an "anticlinal" bend in the fault, γ is a double-valued
function of θ and φ. Folds with larger and smaller values of γ are called first-mode
(Mode-I) (Fig. 13.7a) and second-mode (Mode-II) (Fig. 13.7b) folds, respectively. If the
angular parameters of a fault-related fold are not related to each other through eqs.
13.1–13.3, then it is either not an FBF or assumptions for kinematic modelling are
invalid.

85 0 75 0 45 −4 −5 0 −65 0 −75 0
65 0 55
90 0 50 90 0
0 0
5 0 5

β = −5 0
50
25

−25
12 0 Mo 15
35

−3
0 d
0
0

Mo - I e

5
0
50
15

−15
d e-
II
0

30 0

0
15 0
45 0
10 0
φ=

0 30 0
1
0

60 0 60 0
Axial angle, γ

60 0
45 0
75 0
60 0
β= φ=90 0
14 0 0 75 0
φ=
90 0
γγ 30 0
0
30 φ β
γ γ
β
Anticline φ
θ φ=θ −50
θ
Syncline
00 00
0 0 0 0 0 0
90 60 30 0 - 60 - 30 - 900
Initial cut-off angle, θ (anticlines) Initial cut-off angle, θ (synclines)

Figure 13.6. Graph showing relationships (equations 13.1-13.3) between


different angular parameters of fault-bend folds (after Suppe 1983). The
left and right sides of the graph are for anticlinal bend and synclinal
bend, respectively, in the fault. The graph shows that initial cut-off angle
for anticlinal bend cannot be more than about 30° for a simple step up in
décollement (φ = θ). Also γ will have two values for every value of φ, for
anticlinal bend in the fault; fold with larger and smaller γ values are
called Mode-I and Mode-II folds, respectively.

a b
Mode-I fault-bend fold (larger axial angle) Mode-II fault-bend fold (smaller axial angle)

Figure 13.7. Mode-I and Mode-II fault bend folds.


128
13.4 Fault-propagation folding

A fault may not propagate rapidly through rock sequence as a clean fracture but
may propagate gradually as slip accumulates. In such a case, at each instant during
fault propagation, slip decreases upsection to zero at the fault tip and the shortening is
transferred to a fold developing above the fault's tip. This kinematic process is called
fault-propagation folding (Suppe and Medwedeff 1984, 1990; Suppe 1985; Mitra 1990).
The folds formed above the tip of the propagating fault are called fault-propagation
folds (Figs. 13.8). The primary geometric assumptions are the same as listed in section
13.1. The relations between the step-up or footwall cut-off angle (θ), interlimb half-
angles of faulted (γ*) and unfaulted (γ) units, and dip of the forelimb (δ) are as follows
(Fig. 13.8d; Suppe 1985, modified by Mitra 1990):

cot θ + 2 tan (θ/2) = 2 cot γ* - cot 2γ* (13.4)


γ = γ* + (θ/2) (13.5)
δ = 180 - (2γ* + θ) (13.56)

It follows that for a given θ, the geometry (i.e., interlimb angle and limb dip) of the
nascent fold is maintained throughout its history. In foreland fold-thrust belts, θ
usually varies between 15°-30°; the corresponding values for γ* and δ are 21.6°-38.8°
and 58.1° (overturned)-72.4° respectively. Therefore, in an ideal model, the fault-
propagation folds are usually asymmetric and tight with steep to overturned forelimb.
The ramp anticline is sharp-hinged with only one axial plane (AB', Fig. 13.8), up to the
plane that locates the fault tip, i.e., the contact between faulted and unfaulted layers.
Beyond this plane the axial surface bifurcates (axial planes A and B’) and the anticline
is flat crested. The axial planes of the leading and trailing synclines terminate at the
fault tip and at the fault bend, respectively. As the fault tip climbs up through the
section with increasing slip, the axial plane AB’ grows in length as the width of the flat
crest decreases.

129
A B'

A B'
A' B A' B

AB' slip

a fault tip AB' c


fault/fold initiation point
A B'
δ γ γ
A' B

γ∗ γ∗ unfaulted layer
AB'
faulted layer
b d θ

Figure 13.8. Progressive development of fault-propagation fold (after


Mitra 1990; Medwedeff and Suppe 1990). Angular parameters used for
kinematic modelling are shown in (d).

13.5 Modified models of fault-bend and fault-propagation folding

The theories of fault-bend and fault-propagation folding are powerful end-member


models, which can be modified to produce a variety of complex fold shapes. A few
examples of such modifications are described below:

• Breakthrough structures in fault-propagation folding: In the fault-propagation


folding model, the fault may propagate self similarly all the way to the surface
or it can be halted at any instant depending on the rock properties. In the latter
case, the folding ceases and the fault may break through in a fracture mode. The
"breakthrough" thrust may propagate in several ways, e.g., along a décollement
surface, synclinal axial surface, steep forelimb of the anticline and anticlinal
axial surface (Mitra 1990; Suppe and Medwedeff 1990). Few examples of fault-
propagation breakthrough structures are shown in Fig. 13.9.

slip slip
slip

Decollement breakthrough Anticlinal breakthrough Synclinal breakthrough


(a) (b) (c)

Figure 13.9. Examples of breakthrough structures associated with fault-


propagation folding (after Suppe and Medwedeff 1990)

130
• Multi-bend fault-bend folding: If a thrust has sufficiently large slip, the beds may
slip past more than one bend in the fault producing "multi-bend fault-bend"
folds (Suppe 1983; Medwedeff and Suppe 1997). Examples of multi-bend fault-
bend folding with two bends in the ramp portion are shown in Fig. 13.10.
Medwedeff and Suppe (1997) show that multiple fault bends give rise to
complex fold shapes by a combination of two processes: a process of kink-band
interference, and a set of processes associated with the generation of new dip
panels and axial surfaces as hangingwall cut-offs are displaced past successive
fault bends in the footwall. In theory, curved faults can be approximated by an
arbitrary number of straight segments. In practice, a small number of straight
segments generate a high degree of complexity and adequately models fold
geometry. Consequently, a curved ramp can be modeled as a quasi-curved
ramp. Also multi-segment ramps lead to proliferation of non-parallel axial
surfaces that produce quasi-curved fold shapes.

Bend 3 Bend 3
Bend 2
Bend 2
a b
Bend 1 Bend 1

Figure 13.10. Multi-bend fault-bend fold (after Medwedeff and Suppe


1997). (a) Synclinal bend. (b) Anticlinal bend.

• Simple shear in the faulted layers: In the ideal models of fault-related folding there
is no layer-parallel simple shear within the hangingwall block. Consequently,
the beds do not undergo layer-parallel shear until they enter the fault-bend fold
and the fault surface is always the "active slip surface". If this condition is
relaxed, i.e., if layer-parallel simple shear within the thrust sheet is allowed,
then the fold shape can be modified in many different ways. The ideal
theoretical shape of a fault-bend fold associated with a simple step in
décollement is a flat-crested anticline (Fig. 13.5). Suppe (1983) shows that if the
thrust sheet undergoes pervasive layer parallel simple shear, the two axial
surfaces of the flat-crested anticline progressively annihilate each other forming
131
a new axial surface resulting in a sharp-crested fold (Fig. 13.11). The
annihilation involves locking of the fault surface and migration of the active slip
surface progressively to higher bedding surfaces, resulting in a simple shear in
the hangingwall of the thrust sheet above the lower décollement. The active slip
surface is always in the bed in contact with the branch in the axial surface. All
the slip is absorbed in the annihilation and the thrust sheet is immobile along
the lower décollement, beyond the anticlinal axial surface.

Ac
tiv es
lip
su r
Ac fa c
tiv e
es
lip
su r
fa c
e
α

a b

Figure 13.11. Development of sharp-crested anticlinal fault-bend fold


through annihilation of flat-crest. Annihilation takes place as active slip
surface migrates up section from the ramp (After Suppe 1983).

• Combined fault-propagation and fault-bend folding: In the model of fault-


propagation folding, the folding initiates as soon as the ramp begins to step up
from the décollement (Fig. 13.12). The fold acquires its basic geometry at this
stage and continues to grow self similarly with the propagation of the fault.
Also all the beds in the hangingwall cut by the fault are folded through the
anticlinal axial surface. Chester and Chester (1990) made an interesting
modification to this model wherein they suggest the existence of a pre-existing
ramp (or fracture) (Fig. 13.12). Folding is initiated when the pre-existing ramp is
activated without any change in dip. In a way, it is similar to fault-bend folding
where fracture forms first followed by folding. The difference is that in Chester
and Chester's (1990) model there is no upper flat and the ramp continues to
propagate without change in orientation. The fold above the fault tip is a fault-
propagation fold whereas the fold at the ramp-flat intersection is a fault-bend
fold. These two folds are separated by an unfolded region. Another important

132
geometric difference is that in this model some of the lower layers in the
hangingwall cut by the fault are not folded through the anticlinal axial surface.

fault tip

a b
fault/fold initiation point
Fault-propagation fold
No deformation
Fault-bend fold

slip

c d

Figure 13.12. Progressive development of a fault-related fold, which


combines fault-propagation and fault-bend folding models (after
Chester and Chester 1990).

• Forelimb thinning/thickening: Jamison (1987) noted that interlimb angle and


forelimb dip in many natural fault-bend or fault-propagation folds are different
from values that are predicted from simple models with constant layer
thickness. Forelimb thinning/thickening was suggested to be a solution to this
problem (Fig. 13.13).

residual

a b

Figure 13.13. Forelimb thinning and thickening model of Jamison (1987)

Forelimb thickening leads to larger interlimb angle and smaller forelimb dip than
the predicted values. For a given value of θ, forelimb thinning leads to folds with
smaller interlimb angle and larger forelimb dip but forelimb thickening leads to folds
with larger interlimb angle and smaller forelimb dip. The thickening/thinning occurs
only in the forelimb, the thickness remains constant in the remainder part of the beds.

133
Also, the basic geometry of the fold is acquired at the time of inception. However, the
shape of a fold-propagation fold can be modified in the case of a décollement
breakthrough; the part of the fold that retains the original geometry is called the
residual. Jamison (1987) also demonstrated that simple shear parallel to the thrust sheet
thins the forelimb and reduces the interlimb angle and thus changes the fold shape. In
their detailed modelling of fault-propagation folding, Suppe and Medwedeff (1990)
developed a theory, called the "fixed front anticlinal axial surface" theory, in which
forelimb thinning/thickening was considered as a possible variable.

13.6 Décollement fold

Décollement folds (also called detachment folds) are the third end member of fault-
related folds (Fig. 13.14).

Lift-up fold
Box fold

slip slip

(a) (b)

Footwall syncline Footwall syncline +


breakthrough

slip slip

(c) fault tip (d)

Figure 13.14. Décollement (or detachment) folds (after Mitra and


Namson 1989; McNaught and Mitra 1993).

These folds develop in response to shortening above a décollement surface or a


thrust that is parallel to bedding. These folds are not associated directly with ramps.
They require a ductile décollement layer (e.g., salt or shale) that can infill the space
generated at the base of the fold. Décollement folds are rootless and commonly
disharmonic. Box folds (Fig. 13.14a) and lift-off box folds (Fig. 13.14b) are two types of
décollement folds (Mitra and Namson 1989). There can be breakthrough associated

134
with these folds that can leave a syncline, called footwall syncline (McNaught and Mitra
1993) stranded in the footwall (Fig. 5.12a,b).

13.7 Rollover anticline

Rollover or rollover anticline (Hamblin 1965; Gibbs 1984) is folding in the


hangingwall block as consequence of slip on a listric normal fault. Such folds are
widespread in extensional tectonic set up. Kinematic modelling of folds above a curved
fault is difficult. However, we can model rollover on normal fault with one sharp bend.
This can be followed by modelling with multiple bends that approximate a listric fault.
Fig. 13.15 shows a model of rollover development for a single sharp bend on a
normal fault (Xiao and Suppe 1992). In this model a fracture with a bend forms through
the rock sequence (Fig.13.15a) and before the hangingwall starts to move. Therefore,
this model is similar to the fault-bend fold described in section 13.3. As the
hangingwall was pulled away, a void is created (Fig. 13.15b). The hangingwall
collapses to fill the void and a kink band bound two axial planes form (Fig. 13.15c). It is
assumed that the orientation of the collapse surface will be given by coulomb failure
criteria (see section 7). The upper axial plane is called the active axial plane because it is
the locus of active folding and the material points roll through this axial plane. The
active axial plane remains fixed to the footwall at the bend. The inactive axial plane is
anchored to hangingwall block and is located at the boundary between deformed and
undeformed rocks. In the next step, the hangingwall moves and collapses to fill the
void increasing the width of the kink band (Figs. 13.15d, e). In reality the movement
and collapse are not stop-and-go type as in the model. The collapse and widening of
the kink band is continuous with increase in slip.
Normal faulting decreases elevation on the hangingwall side of the fault leading to
the formation of basin. The basin may be filled up with syntectonic sediments, which
are called growth strata. The geometry of rollover gets suitably modified by growth
strata (Fig. 13.16). A detailed description of rollover development during active
sedimentation is give by Xiao and Suppe (1992).

135
Active axial Inactive axial
surface surface

(a) (c)
Coulomb failure
surface
0
70

Void (d)
(b)

Active axial
surface Inactive axial
surface

(e)

Figure 13.15. Kinematic model of rollover development (after Xiao and


Suppe 1992).

Growth wedge
(a)
Growth
Zone of most strata
Pre-growth
recent deformation
strata
Growth axial surface

(c)

(b) Active axial Inactive axial


surface surface

Figure 13.16. Progressive development of rollover with growth


sedimentation (after Xiao and Suppe 1992).

136
137
14. Balanced Cross Section: Introduction
14.1 Basic philosophy

The whole purpose of balancing structural cross sections is to limit the interpretation
to that which can be considered to be structurally reasonable. Two interpretations of a
faulted anticline are shown in Fig. 14.1. The orientations of the faults suggest that they
are conjugate faults. The structural interpretation in Fig. 14.1a is unreasonable because
the two faults have normal and reverse senses of displacement but we know that
conjugate faults should have same sense of displacement. Further, the layer thickness
varies in a way for which there is no obvious explanation. But in Fig. 14.1b both the
faults have normal sense of displacement and the fold follows parallel (class 1B)
geometry. Thus the interpretation in Fig. 14.1b is reasonable. Most of the published
cross sections are schematic and many of them are reasonable. A cross section across
frontal fold-thrust belt of NW Himalayas is shown in Fig. 14.2. Although the section is
schematic yet it is reasonable in the sense that the interpretation is in conformity with
structures seen in other fold-thrust belts. However, we should avoid constructing
schematic sections based on imagination or intuition that may lead to structurally
unreasonable or untenable interpretation. It is desirable that geometrically and
kinematically valid constraints should be applied during and after section construction
so that our subsurface structural interpretation is reasonable in terms of present
knowledge about how rocks deform. A balanced cross section is simply a better cross
section as compared to a free-hand drawn schematic and unbalanced cross section. It is
important to remember that a balanced cross section is constructed using limited data
set and a few reasonable assumptions. Techniques of cross-section balancing are
described in Dahlstrom (1969), Suppe (1985), Ramsay and Huber (1987), Marshak and

138
Woodward (1988), Woodward et al (1989), Mitra and Namson (1989), Mitra (1992) and
others.

(a) (b)

Figure 14.1. A simple example of “unreasonable” (a) and “reasonable”


structural interpretation. Interpretation in (a) is “unreasonable” because the
conjugate faults have opposite senses of movement and layer thickness
variation is unexplainable. But in (b) both the faults have normal sense of
movement and the fold has class 1B geometry.

r hi
pu uk r
opi am r pu
ag t l ro st a m st BT NE
SW an st er s wa st Pathru
l
Pathru
FT So thru D thru Ja thru M
M

20 km

Siwalik Subathu-Dharmsala Lesser Himaya Zone Metamorphic basement

Figure 14.2. A schematic yet reasonable cross section across Himalayan fold-
thrust belt, Kangra area, Himachal Pradesh (Ranga Rao 1989, unpublished
ONGC report, in Biswas 1994). Although the section is not balanced, the
interpretation follows structural styles in fold-thrust belts.

14.2 Definition of balanced cross section

Dahlstrom (1969) formally introduced the concept of balanced cross section in the
literature and suggested two rules to be followed while constructing a balanced cross
section:

• Bed lengths must be preserved during deformation. In other words, deformed


and undeformed bed lengths must match or “balance”.
• There are only a limited suite of structures that can exist in a specific geological
environment. The “foothills family of structures” in the Canadian foothills

139
comprise of concentric folds, décollement, thrusts (usually low angle and often
folded), tear faults, and late normal faults.

Elliot (1983) gave a more restrictive definition of a balanced cross section:

• A balanced cross section must be both admissible and viable.


• The structures drawn on a section should be those that can be seen in the field
in outcrops, cliffs, mountain sides etc.
• The use of these structures leads to an admissible cross section.
• If a section can be restored (i.e., retrodeformed) to an unstrained state, it is a
viable cross section.

There are areas where exposures are scanty owing to heavy vegetations, such as
Himalayan foreland belt. In such areas it may not be possible to decide admissible
structures as suggested by Elliot (1983). Therefore, a preferred definition is:

• A balanced cross section is a geometrically correct (i.e. admissible) deformed-state cross


section, which can be restored (i.e., viable) to an undeformed state or to an earlier less
deformed state through kinematically valid steps.

14.3 Two aspects of balanced cross sections

From the foregoing discussion, it is apparent that there are two aspects of balancing
a structural cross section:

• Construction of a valid structural cross section, which depicts the present day sub-
surface structural geometry. This is the so-called deformed-state cross section,
which geologists have been constructing through ages. The key issue here is
that the section we construct must contain the geometry of structures we
actually see in the field or the geometry of structures present in other areas with
similar tectonic setup. For example, if we are to construct a cross section across
a rift basin, the section must dominantly contain normal faults and associated

140
structures. Or, if our section is across a fold-thrust belt, the main structural
architecture must be controlled by thrusts and associated structures. This
exercise results in admissible deformed-state cross section.
• The deformed-state cross section should be restored to undeformed state (i.e., validated).
Cross sections are essentially interpretations based on incomplete data and
require extrapolation and interpolation. So, how do we increase confidence in
our interpretation? Simply put, we validate. If we cannot validate a cross
section, we reject the interpretation. We then reinterpret the available data and
construct a new section and try to validate. A balanced cross section is the one,
which can be validated. We know that sediments are commonly deposited with
(sub-) horizontal layering. If we undeform (i.e., restore or retrodeform) an
admissible cross section drawn across a deformed sedimentary basin and find
that the layers become horizontal then we have validated our cross section.

We have not said anything about a “balanced cross section” being a correct
interpretation. Balancing a cross section is akin to inverse modelling. Like all inverse
modelling, the solution we arrive at is non-unique. It is possible to construct more than
one balanced sections with the same data set. Therefore, it is important to remember that
balancing is a necessary but not a sufficient criterion for a correct structural interpretation.
Balancing even a mildly complex structural section is a time consuming, tedious and
frustrating exercise without any guarantee of success within the fixed timeframe of a
project. However, while non-balanced and schematic sections are almost always
wrong, a balanced cross section is probably correct! However, practicing the techniques
of cross-section balancing, particularly in deformed sedimentary basins, should lead to
better and more reliable cross sections, even if a section cannot be rigorously balanced.

14.4 Examples

A few examples will illustrate the requirements for balanced cross section. The
hypothetical deformed-state cross section in Fig. 14.3a seems to be quite reasonable
with thrusts and related ramp anticlines. The steeply dipping parts of the thrusts may

141
be explained by rotation caused by younger faults in the footwall. However, if the
section is restored (Fig. 14.3b) keeping the bed length and sectional area constant, the
thrust trajectories become unreasonable. Therefore, the deformed-state cross section is
not a balanced cross section and should be discarded. The hypothetical section in Fig.
14.4a is a balanced cross section because it can be retrodeformed into an admissible
restored section. Fig. 14.5a shows a section across NW Himalayan foreland fold thrust
belt, Dehra Dun (Mishra and Mukhopadhyay 2002). The Mohand anticline is a multi-
bend fault-bend fold with about 12% foreland thinning. A multi-bend fault-
propagation model with about 47% forelimb thinning is appropriate for the
Santaurgarh anticline. The Santaurgarh anticline is related to a blind thrust with
anticlinal breakthrough structure (Santaurgarh thrust). The admissible restored section
of the deformed-state section is given in Fig. 14.5b. Therefore, the cross section in Fig.
14.5 is a balanced cross section. It is emphasized that both deformed and restored
section should be given because unless an admissible restored section is given,
balancing is not demonstrated.

Deformed

(a)
Restored

(b)

Figure 14.3. The hypothetical cross section in (a) is unacceptable because the
restored section is not admissible.

Deformed

(a)
Restored

(b)

Figure 14.4. The hypothetical section in (a) is a balanced cross section because
both the deformed-state section and restored sections are admissible.

142
Santaurgarh Thrust
Santaurgarh anticline
Main Frontal Mohand anticline Main Boundary Thrust (MBT)
Thrust (MFT)
Doon Gravel Topography
P Alluvium L Doon Gravels/
Alluvium
Up. Siwalik Fm
? Mid. Siwalik Fm
Lr. Siwalik Fm
Dharmsala Gr
(a)
Subathu Gr
Detachment
L
P
MFT

?
(b) 5 km

Figure 14.5. Balanced (a) and restored (b) cross sections across Himalayan
foreland fold-thrust-belt, Dehra Dun area (after Mishra and Mukhopadhyay
2002).

14.2 Types of cross sections

There are different types of cross sections and the subtle differences between them
should be understood clearly:

• Deformed-state cross section: Any structural cross section showing the present-
day sub-surface structural geometry.
• Admissible cross-section: Deformed-state cross section in which the depicted
structures conform to the available data, and are also geometrically and
kinematically admissible.
• Restored cross section: A cross section that has been 'pulled apart' in the sense
that displacements on faults have been removed and folded beds have been
straightened.
• Admissible restored section: It is a restored section in which the restored beds are
horizontal and the restored fault trajectories are admissible. In particular, ramp
dips should not be more than 30-35° and thrust trajectories should not have zig-
zag shape unless they are out-of-sequence thrusts.
• Viable cross section: A deformed-state cross section that can be restored to an
admissible restored section.

143
• Balanced cross section: A deformed-state cross section that is both admissible and
viable. Therefore, a balanced cross section has been restored to undeformed
state and has been tested for viability.

14.1 Fold-thrust belts (FTB)

Fold-thrust belts (FTBs) form when sedimentary basins are deformed by


compressional tectonics. Much of the concepts of balanced cross sections have emerged
from fold-thrust belts. Consequently, section construction and restoration are described
with reference to fold-thrust belts (sections 15 and 16). But these techniques are equally
applicable for extensional basins. The characteristic features of fold-thrust belts include:

• Narrow and elongated basins and occur at the margin of orogenic belts.
• Composed mostly of sedimentary rocks.
• Deformed in response to compressional tectonics.
• Deformation is partitioned mostly in the cover rocks and basement remains
largely unaffected – the so-called thin-skinned tectonics (Fig. 14.6). The cover
sequence and basement are separated by a major thrust of décollement.

Hinterland Foreland

SW NE

Basement Regional detachment 20 km


(a)

Hinterland Foreland
NW SE
mean sea level

20 km
semen t
crystalline ba
Pr ecambrian
Basal detachment
(b)

Figure 14.6. Cross sections across foothills of (a) Canadian Rockies (after Price
1981) and (b) NW Himalayas (after Mukhopadhyay and Mishra 1999) showing
essential feature of thin-skinned tectonics where cover sequences are deformed
but basement remain largely unaffected.

144
• Presence of hinterland-foreland pair (Fig. 14.7). The undeformed sedimentary
basin in front of the FTB is called foreland and the metamorphic core with
ductile deformation is called hinterland. The external zone contains the FTB. The
terms external zone and foreland are sometimes used interchangeably.
• Folding is commonly a consequence of thrusting.
• Axial and fault traces are parallel to each other and parallel to the trend of the
orogenic belt.

High Himalaya sedimentary Zone


N High Himalaya Crystalline Zone
Lesser Himalaya Zone
TR Sub-Himalaya Zone
AN
H
S ITSZ: Indus Tsangpo Suture Zone
IN STDS: South Tibet detachment Sysytem
EX T MCT: Main central Thrust
E
TE R MBT: Main Boundary Thrust
LA
RN N MFT: Main Frontal Thrust
FO A D
RE L
LA Z
O
HI M
0 N N ALA
30 D E YA
ITSZ
Delhi MB
Te T MC
tr a cto n T STDS
nsp ic
ort
250 km AL L U MF
VIA L T
80
0
PLAIN 900

Figure 14.7. Geological map of the NW Himalayas (simplified after Gansser


1981) showing hinterland, foreland, external zone and tectonic transport
direction.

The frontal part of the Himalayas shows all the characters of a fold-thrust belt. The
Sub-Himalaya Zone consisting of Tertiary sedimentary rock sequence and the
Precambrian sedimentary and very low-grade metamorphic rocks of the Lesser
Himalaya Zone together constitute the Himalayan FTB (Fig. 14.7).

14.6 Applicability

The concepts and techniques of balancing cross sections originated in foreland fold-
thrust belts, and in particular in the frontal fold-thrust belt of the eastern Canadian
Rocky Mountains, which are characterized by folded and faulted, non-metamorphosed
sedimentary sequences that lie above a gently hinterland-dipping detachment or

145
décollement. These concepts and techniques, sometimes referred to as Rocky Mountain
Principles, have been successfully applied in many foreland fold-thrust belt in different
parts of the world. These techniques can also be applied in extensional terrains. The
most important application of this technique is found in hydrocarbon exploration and
exploitation where reliable subsurface structural interpretation is of paramount
importance for fiscal planning. A balanced cross section across Bolivian fold-thrust belt
is shown in Fig. 14.8a (Baby et al. 1995). The balanced cross section was used to derive
burial and uplift histories. This in turn helped in defining the gas and oil windows for
detailed exploration (Fig. 14.8b). It is not a mere coincidence that structural geologists
working for oil companies authored many of the early publications on balanced cross
section. This technique is also important for academic research because a balanced
cross section helps us to understand how an orogenic front develops and propagates in
time and space, and to estimate orogenic shortening. In turn, these may help us to
understand the geodynamic evolution of an orogenic belt. Therefore, in the techniques
of cross-section balancing there is a convergence of interests of academicians and
explorationists!

SW NE
5
Isiboro
anticline
Late Paleogene-Neogene
Km

0 Jurassic-Cretaceous
Up. Carb.-Lr. perm.
Carboniferous
-5 Dev.-Sil.-Ord.
Cambrian-Precambrian

(a) 10 km

5
Km

0 Oil window
Oil & Gas Zone
Gas window
-5

(b)

Figure 14.8. An example of application of cross-section balancing in


hydrocarbon exploration in Bolivian fold-thrust belt (Baby et al. 1995).
(a) Balanced cross section. (b) Suggested oil-gas windows based on
burial-uplift histories derived from balanced cross section.

146
147
15. Section Construction
Construction of cross sections is one of the most important tasks undertaken by
explorationists and structural geologists. Sub-surface structural interpretations are
communicated through cross sections. Exploration strategies and financial planning for
explorations for hydrocarbon and other natural resources are based to a large extent on
structural cross sections. Therefore, cross sections should be constructed with utmost
care and, as far as possible, they should be quantitative rather than purely qualitative.
The techniques of balanced cross section have been very useful for quantification of
structural interpretations in sedimentary basins deformed due to in compressional
tectonics.

15.1 Strategic Considerations

• Line of section: Inherent in the methods of cross-section balancing is plane strain.


For this reason the line of section must be chosen strictly parallel to the tectonic
transport direction. However, a line of section within ± 10° of the tectonic
transport direction does not result in significant error (Woodward et al. 1989).
The tectonic transport direction can be determined using bow-and-arrow rule, as
illustrated in Fig. 15.1a (Elliott 1976). This direction can also be determined from
stretching lineations or from the orientations of the axes of small-scale folds. If
the fault traces and axial traces are parallel on the map (Fig. 15.1b), then a
section line oriented perpendicular to fault/axial traces is acceptable in most
situations. In most FTBs, the tectonic transport direction is approximately
known.
• Ramp-flat geometry: The thrust faults in fold-thrust faults usually assumed to
have ramp-flat trajectory with sharp bends for ease in section construction.

148
Slight curvature near the bend can be ignored without adding significant error.
Curved faults can be considered to have multiple bends with straight line
trajectories in between bends.

Thrust tip

Thrust
Axial traces

Direction of Direction of
tectonic transport tectonic transport

(a) (b)

Figure 15.1. (a) “Bow-and-arrow” rule to determine tectonic transport direction,


which is given by the perpendicular to a line which connects the two exposures
of a thrust tip line in map view and is in the direction from tip connector to the
fault (Elliot 1976). (b) The direction perpendicular to the fault and axial traces
also gives the tectonic transport direction.

• Depth to Detachment: When we construct deformed-state cross section in fold-


thrust belts, what we essentially try to do is fill up the space between the
topographic surface and the basal detachment! So it is important to know the
depth and dip of this basal detachment. It may be known from seismic
reflection profiles or borehole litholog data. In the absence of such data, a
commonly used method is depth-to-detachment calculation, which assumes
area conservation during deformation. The method of depth-to-detachment
calculation is illustrated in Fig. 15.2.

1 2 3 4 5

S = lo - l
2' 3' Area A
Area B
1' 4' 5'
d

Shortening, S = lo - l Area A = Area B = d x S d = area A/S

Figure 15.2. Depth-to-detachment calculation.

149
• Reference lines: There are two types of reference lines, viz., pin line and loose line
(Fig. 15.3). A pin line is a line in the deformed-state cross section from which all
restoration measurements are made. Pin lines can be regional or local. Regional
pin lines are perpendicular to stratification and are located in the undeformed
foreland. Local pin lines are located within the thrust belt. Loose lines are usually
marked near the trailing edge of the cross section. A loose line can be
considered to be chain of marker points in the layered sequence. It is
particularly useful in tracking layer-parallel simple shear, which may not be
obvious in deformed-state cross section.

Local pin lines Foreland


Regional
Loose line pin line

Figure 15.3. Different types of pine lines.

• Template: A restoration template shows the assumed pre-tectonic attitudes of the


rocks shown in the deformed-state cross section. Restoration is usually done on
this template. Template can be constructed if stratigraphy of the area is well
known.
• Sequence of thrusting: It used to be assumed that thrusts in fold-thrust belts form
in a forward-breaking sequence. But now we know that break-back sequence,
out-of-sequence (Morley 1988) and synchronous thrusting could be important.
Knowledge of sequence of thrusting is particularly useful for restoration. Faults
should be restored in the reverse order than they formed.

150
• Slip on a fault: If the slip on a fault surface is assumed to be constant then
matching hangingwall and footwall cut-offs is simple. However, for a blind
thrust this is not true. The slip should become zero at the fault tip and the
shortening is transferred to the overlying fold.
• Balanced forward modelling: It should be an integral part of section construction.
As many balanced forward models as possible should be constructed in the
vicinity of each fault. Then, we can attempt to fit these models like a jig-saw
puzzle. Sections that contain forward modelled structures have better chance of
being restorable.
• Previous sections: Do not throw away previous sections, if available. Modify such
sections after proper evaluation in terms of admissibility and viability. One can
avoid repeating the same mistake made by previous workers. Of course if there
is no previous section available one has to start from scratch with raw data.
• Data: All available data should be gathered and compiled, such as surface
geological map, stratigraphy, regional geology and tectonic setting, dip data
collected at the surface, dipmeter data from borehole logs, lithologs, seismic
reflection profiles etc. Obviously more data we have more confidence we will
have in our cross section. Subsurface data, such as, seismic reflection profiles,
are not absolutely essential, for balancing a cross section, which can be done from
geological map and surface dip data alone. Of course, for an offshore project or
if the area is covered by alluvium the first data is commonly seismic reflection
profile. It should be remember that even with today's improved data acquisition
and processing techniques, seismic data often leave much to be desired.

15.4 Parallel Folds

In some geological environment, notably in sedimentary sequences, the folds are


usually parallel, i.e., they belong to class 1B. In such folds, layers do not show any
appreciable thinning of limbs or thickening of hinges, the orthogonal thickness
measured perpendicular to layering remains constant at all points around the fold. In
multi-layered sequence, the folds are conformable or harmonic. The folded surfaces are
151
smoothly curved or concentric in some parallel folds, whereas in others angular or
sharp hinge zones separate domains in which layers are nearly straight. These folds are
called concentric and kink folds, respectively. An interesting feature of kink folds is
that they are parallel as well as similar (class 2).
Parallel folds can be extrapolated data to depth using either Busk method (Fig. 15.4)
or kink method (Fig. 15.5). The Busk method of section construction is appropriate for
concentric parallel folds, whereas the kink method is appropriate for parallel folds with
angular hinge zones and straight limbs. These methods produce reliable cross sections
only if the assumption that the folds are parallel is valid. For non-parallel folding we
should use dip-isogon method or one of the several orthographic projection methods to
construct sections. The Busk and kink methods are described here because they,
particularly kink method, are used most frequently in fold-thrust belt.
Any kind of dip data can be extrapolated to depth using either Busk or kink
method. The dip data can be in the forms of outcrop dips, formation tops or bottoms
taken from outcrop or well log information, or dip meter data. However, it is important
to be consistent.

Busking

The Busk method (Fig. 15.4) allows us to construct the traces of bedding planes in a
section plane from surface or subsurface measurements of the attitudes of the folded
layers. The geometric basis of this method is the assumption that the folded layers are
everywhere tangent to circular arcs. A consequence of this assumption is that the trace
of each folded layer in a profile can be divided into a number of segments each of
which is either a portion of a circular arc or straight line. Along each circular arc
segment, the dip changes smoothly and continuously. Adjacent circular arcs are
connected by inflexion points or by straight-line segments.
Lines perpendicular to dips are drawn from the position of dip measurement data.
The perpendicular lines from two adjacent dip data intersect at a point that represents a
radius for a curvature of an arc, which is utilized to project the beds between the two
data points. The method is described in Fig. 15.4. For example, in Fig. 15.4 two dip data

152
(1,2) are shown at two locations (A, B). We draw perpendicular (C, D) to dips 1 and 2,
the perpendiculars intersect at O. We then draw to circular segments with OA and OB
as radii. This gives us the fold segment between C and D. One of the problem of
Busking is that, singularities (points of infinite curvature) often appear in Busk
construction of folds. Singularities are rarely observed in natural concentric parallel
folds.

G G
B C B C
A A
F J H
D D F

E E

O O

Figure 15.4. Busk method of extrapolation of surface dip data to depth.

Kinking

Kink or constant dip-domain method (Fig. 15.5) has proven to be extremely useful for
extrapolating data to depth. Many fold-thrust belts contain folds, which display a kink
or dip-domain geometry. Therefore this method has become very popular for
constructing cross sections in such belts. Angular folds produce domainal dip
patterns on maps. A dip domain is an area in which strata have nearly constant dip or
dip varies within a small range. Adjacent dip domains are separated by narrow zones,
representing hinge zones, in which dips change rapidly.
In kink methods we use two inherent geometric features of kink folds to extrapolate
surface data to depth: if layers do not change thickness, the axial plane bisects interlimb
angle and only one axial plane extends from the junction where two axial planes meet.
The kink-method of section construction is rather straightforward and the technique is
explained in Fig. 15.5. First we locate boundaries between dip domains (Fig. 15. a) and
draw the axial planes with orientation that bisects the angle given by the dips in two
adjacent dip domains (Fig. 15.5b). Where two axial surfaces meet, a new axial plane

153
starts bisecting the angle between two intersecting axial planes. Within the area
between two axial planes, straight beds are drawn. Round-hinged folds can be
approximated as closely-spaced small kinks (Fig. 15.6).

Domain Boundary
Ground
surface
A B

(a) Dip

A B C D
Domain Boundary
(Axial Surface)

Branch
A B point

Dip Domain X

(b) Dip Domain Y


(c)

Figure 15.5. Kink method of extrapolation of surface dip data to depth.

Figure 15.6. A round-hinged fold can be modelled using kink-fold geometry.

15.5 Drawing a Cross Section

Figure 15.7 illustrates how surface dip data can be extrapolated to depth and sub-
surface fold geometry can be deciphered. The dip data and stratigraphy are given on
the left side of the diagram. At locations A and B contact between sandstone and shale
are exposed. The dip data shows five planar domains (domain 1-5) suggesting kink
fold geometry. At domain boundaries attitude of axial planes are deduced from the
bisectors of the adjacent domain dips. The axial surfaces are extrapolated to depth.
154
Wherever two axial surfaces meet, a new axial surface emerges whose orientation is
given by the bisector of the two axial planes. The beds are then extended using the dip
domain data and the fold is constructed.

Data Solution
Domain 1

Domain 2

Domain 3

Domain 4

Domain 5
75 o 70 o

Well 60 o
30 o 0o 40 o 60 o 0o 40 o

A B A B

Sandstone A Sandstone A
Shale A Shale A
Sandstone B Sandstone B

Shale B Shale B

Sandstone C Sandstone C

Shale C 75 o Shale C

Sandstone D 60 o Sandstone D

Figure 15.7. Extrapolation of surface data to depth using kink method to deduce
the geometry of the large fold.

Maximum
ramp height

Data
Maximum depth
of fault

Solution 1 : Maximum depth of fault Solution 3 : Maximum ramp height

Figure 15.8. A hypothetical example of section construction showing range of


possible solutions. See text for discussion.

155
Fig. 15.8 shows an exercise of how surface dip data and known stratigraphy are
used to deduce buried thrust. The dip data show five dip domains, three horizontal dip
domains are separated by two dip domains where dips are steeper. Overall the fold
geometry is that of a flat-crested anticline, so we guess that it is a fault-bend fold. If this
is the case there has to be a thrust at depth. Let us suppose we know that the tectonic
transport direction is towards left. Note that the axial angle (γ) is known and back limb
dip can be used to infer ramp dip or cut-off angle (θ). With available information we
cannot deduce the exact location of the fault. However, we can find range of possible
solutions. Solution 1 is based on maximum possible depth of the fault and solution 2
gives us maximum ramp height. We recognize that an exact solution is not possible in
this example but the range of possible solution can be useful for planning further
exploration strategies.
The examples shown in Figs. 15.7 and 15.8 are hypothetical. A real-life example of
section construction with limited data is shown in Fig. 15.9 (Suppe 1983). Fig. 15.9a
shows the available data near the crest of the Hokou-Yangmei anticline, Taiwan FTB.
The Well A encountered a double thickness of the distinctive Pliocene Chinsui Shale
and normal thickness of formations below Chinsui Shale, suggesting that the small fold
on which Well A sits does not extend below Chinsui Shale. Two guesses were made, as
shown in Fig. 15.9b, both involving a simple step of a thrust fault from one décollement
to another in the Chinsui Shale. In solution 1, a thrust steps up to the north and in
solution 2 a thrust steps up to the south. The important angular observations are that
the dip at the base of the Chinsui Shale is 5° whereas the minimum dip of the Chinsui
Shale, between two wells is 32°. Therefore, we choose 32° - 5° = 27° as θ = φ in solution
1 and β in solution 2. Using Suppe's (1983) equations (or graph) we obtain = 34° for
solution 1 with 34° - 5° = 29° as the predicted surface dip. This predicted dip is much
greater than the observed surface dip of about 16°, so discard this solution was
discarded. For solution 2 we obtain θ = φ = 22° and 22° - 5° = 17° as the predicted
surface dip, in good agreement with the observation. Therefore solution 2 may be
considered viable. We can now compute how the shallow fault in solution 2 will be
folded by the deeper anticline (γ = 58°). The cross-cutting fault block is in the footwall,

156
convex towards the fault; therefore it is a "syncline" and φ = 57° and β = 15°, which are
in reasonable agreement with surface dips. The final interpretation of the structural
geometry using solution 2 is shown in Fig. 15.9c. In this example, a hypothesis was
invoked (fault-bending over a simple step up of décollement) and tested against the
available data and a solution was found.

Well
B
North South

Thickness dip = 320


-1 doubled Chinsui shale
Km dip = 50
-2

-3 (a) Data

16 0 16 0
β =?
(34 )
0 θ = φ0
= 22
β = 270
φ = 0?
0
32
(57 ) Final interpretation
θ = φ0 0 (c) based on solution 2
γ γ = 58
0
= 32 0−5
= 27 β =?
(15 0)

Solution1 Solution 2
(b) (rejected) (accepted)

Figure 15.9. Actual example of quantitative section construction, Hokou-


Yangmei anticline, Taiwan (Suppe 1983). With available data (a), two solutions
are guessed (b). Solution 1 leads to conflict with surface dip data. Solution 2 is
in conformity with surface dip data. (c) Final interpretation based on solution 2.

A different approach is illustrated in Fig. 15.10, where the trajectory of Main Frontal
Thrust (MFT) and the geometry of the Mohand anticline, Dehra Dun re-entrant have
been constrained (Mishra and Mukhopadhyay 2002). The available data are shown in
Figs. 15.10a,b; the MFT trajectory and the basal detachment were approximately
constrained from published ONGC seismic reflection profile and well data. Several
forward models were made, three of which are shown in Figs. 15.10c-e. A model based
on multi-bend fault-bend folding with 12% forelimb thinning and uniformly tapering

157
layers conforms to the surface dip data and interpreted litholog and "best" explains the
geometry of the Mohand anticline (Fig. 10e).
The above examples show that it is possible to construct quantifiable structural
cross sections.

N Dun Gravels
20
Alluvium Mid. Siwalik Up. Siwalik Dun Gravels
Well
15 MhA Topography
22
30
11
MFT and Basal detachment
30
MFT Approximated from
33
subsurface data
MhA

23 5 km
36
MFT? IU M
5 km AL LUV (a) MhA : Mohand Anticline (b)

(c) 5 km
Multi-bend fault-bend folding model with MFT emergent

(d) 5 km

12% forelimb Topography


thinning

sli p
4.0 km

(e) 5 km
Doon Gravels Up. Siwalik Mid. Siwalik Lr. Siwalik Dharmsala

Figure 15.10. A actual example of section construction, Mohand anticline, Dehra


Dun area Himalayan FTB (Mishra and Mukhopadhyay 2002). (a) and (b)
Available data. (c) Solution assuming multi-bend folding model with MFT
emergent. (d) Solution assuming multi-bend folding model with MFT blind. (e)
Solution assuming multi-bend folding model with two synclinal bends on MFT,
uniformly tapering layers, and 12% forelimb thinning. This section is in
conformity with surface dip data.

158
159
16. Section Restoration
Restoration (or check for viability) of a cross section is the actual process of pulling-
back cross sections in contractional terrains and pushing forward cross sections in
extensional terrains. Restoration is done on the restoration template, which should
incorporate changes in stratigraphic thickness, unconformities and any structures
believed to have been present prior to the deformational event which will be restored.
Sections are sometimes partially restored, which means restoration to an earlier less
deformed state. Regional sections should be fully restored. Except possibly for very
simple structures, restored section should invariably accompany deformed-state
section, if it is claimed that the section is balanced.
There are several reasons for restoring a section: (1) It is the ultimate test of whether
the deformed-state section is reasonable or not. A cross section that does not restore is
most likely a bad section. (2) The restored section gives us the original undeformed
length and comparing it with the length in deformed-state cross section allows us to
estimate total shortening (or extension). (3) The restoration process glaringly reveals
commonly made mistakes and errors in section construction.
Restoration is not merely measuring lengths (or areas) in deformed-state section
and stretching them out, and somehow matching them. Sections must always be
sequentially restored, i.e., one fault should be restored at a time. This requires that we must
always take into account the sequence of faulting. The displacement on a younger fault
should be removed first before the displacement on an older fault is removed. Also,
out-of-sequence thrusts should be restored first followed by restoration of in sequence
thrusts. In other word, restoration of faults should be carried out in an order reverse of
the order in which they form. At every stage of restoration, all the structures must
essentially be admissible. If this not the case the section is not balanced even if the
section can be restored to an undeformed state. Step-wise restoration has two
160
additional benefits: (1) it helps us understand how a fold-thrust belt evolves through
space and relative time, and (2) burial and uplift history can be worked out that may be
useful in understanding source-rock maturation and hydrocarbon migration and
accumulation.
The iterative process involved in restoring a section to its undeformed state and
then changing the original section if it does not restore can be tedious, frustrating and
very time consuming. Therefore, it is important to spend the maximum time and effort
to correctly construct the first section using forward-modelled structural geometry.
This is because even minor changes in one part of the section will invariably result in
changing the remainder of the section. The ultimate objective is to construct a
geometrically reasonable cross section within a limited amount of time.
There are two methods of restoration: equal line-length restoration (also called
sinuous bed method) and equal-area restoration. The methods are briefly discussed
below.

16.1 Equal line-length restoration

This is the most commonly applied restoration method. Line-length restoration is


used when it is believed that there has been no stratigraphic thickness changes as a
result of deformation, in any of the stratigraphic horizons included in the cross section.
In other words, the line lengths (i.e., contacts between stratigraphic units in 2D) remain
constant through the deformation. We can restore sections merely by stretching out the
lines to return them to regional level and dip. If thicknesses do not change during
deformation, areas will also remain constant. Consequently, we do not need to
compare areas of beds or thrust sheets. If the folds are kink style, bed lengths can be
measured with a divider or a good-quality ruler. If the folds are concentric, the
measurements obviously become more involved.
Before we start restoration a pin line and a loose line at either end of the section are
to be established. The first step in the restoration is to match hangingwall cut-offs to
respective footwall cut-offs, starting with the frontal thrust sheet if there is no out-of-
sequence thrusting The measured bed lengths are then laid out as straight-line

161
segments. The ends of these straight-line segments then define the footwall trajectory of
the second fault. This procedure is repeated successively for all the faults.
Fig. 16.1 shows a fault-bend fold. For the purpose of restoration two reference lines
are chosen. The pin line and the loose lines are located in the leading and trailing
edges, respectively. Point A was at the upper bend (A') before deformation, so A is
pulled back to this point. In so doing, we also pull back the entire rock package as well
as the reference lines. Now keeping the pin line fixed, we straighten all the lines
keeping the length of the lines constant. In Fig. 16.2, there are two faults in the
deformed section; fault 1 is younger and fault 2 is older. We first restore fault 1 and
then restore fault 2 following the same procedure as in Fig. 16.1. Note that the dip of
fault 2 has changed in restored section. One surprise is that we find significant layer-
parallel simple shear in the restored section that is not obvious in the deformed-state
section.

Deformed Restored
Loose line Pin line

3 4 1' 2' 3' 4' 5' 6'


1 2 5 6

A A'

Figure 16.1. The line-length method of section restoration involving one fault.

Deformed Restored
Loose line Pin line
L2 L1
Shear L2 L1
1
2
ψ 2 1
θ2 θ1 θ3 θ1

Figure 16.2. The line-length method of section restoration involving two faults.

16.2 Equal area restoration

Area restoration is used in regions where it is believed that changes in bedding


thickness have occurred as result of deformation. The changes in bedding thickness

162
must be the result of plane strain within the cross sectional plane in order for the area
balancing to be valid. The changes in bedding thickness must not be result of material
moving in and out the plane of section during deformation. Under this condition, if
there has been no overall volume change then the cross-sectional area of any rock unit
shown in the deformed-state cross section has not changed during deformation.
However, line-lengths (e.g., distances between given thrust faults as measured along
specific bedding planes) may have changed. In such a situation, areas of rocks and not
line lengths are measured for the purpose of restoration.

3
2 4

X 1 A = 3.2 sq. km
Y

X' 1' 2' 3' 4' Y'


A = 2.0 sq. km

A = 3.2 sq. km

Figure 16.3. Equal-area method of restoration (adapted after Marshak and


Woodward, 1988).

Equal-area restoration involves measurement of area (A) of the deformed thrust


sheet and the original thickness (t) of the unit. Assuming plane strain, the restored
average length (L) is given by A/t. Fig. 16.3 shows general methodology of area
restoration. The shaded unit in the rather unusual-looking ramp anticline has
undergone extreme thickening and has an area of 3.2 km2. We take the thickness at the
trailing edge as the original thickness. The line-length restoration of the shaded unit
results in an area of 2.0 km2, which is about 38% too small. In order to keep the area
constant, the undeformed length of the unit must be increased as shown in the lower
diagram.

163
16.3 The key-bed method

Mitra and Namson (1989) pointed out that area restoration does not ensure that the
units are balanced or the fault trajectories in the restored section are geologically
reasonable. This is because only one average length is calculated for each unit and the
top and bottom of the unit are assumed to be equal to the average length. This makes
the undeformed sheet to be a parallelogram.

Key bed
A1
X' A2 L2k L1k
Y'

(a)
Deformed section W' Z'

X Y
A2 A1
L2a L1a
(b)
Area restoration W Z

X L2k Y* L1k

A2 A1
L2a L1a
(c) Ld
L2 Z* L1
Area and key-bed restoration W

Figure 16.4. Combined equal-area and key-bed method of section restoration


(after Mitra and Namson 1989).

This method is illustrated in Fig. 16.4. The deformed-state section is shown in Fig.
16.4a. Equal-area restoration (Fig. 16.4b) results in two parallelogram shaped restored
thrust sheets. The orientation of the thrust YZ and the right-side reference line can be
varied considerably maintaining a constant area, as shown by dashed lines. The equal-
area restoration method can be improved by combining it with key-bed method of

164
restoring individual bed lengths. The key-bed method identifies a thin competent key
unit that undergoes minimum penetrative deformation and area change so that it can
be restored using line-length method. The top layer is assumed to be key bed whose
length remained unchanged. Fig. 16.3c shows combined key-bed and equal-area
restoration. Note the back shear in restored diagram which is not obvious in the
deformed section.

16.4 EVALUATING AND IMPROVING A CROSS SECTION

Restoration of a cross section allows us to evaluate whether our interpretation, i.e.,


the deformed-state cross section, is reasonable or not. In the restored section, trajectories
of all the faults must be admissible in the sense that they should be gently to moderately
dipping towards hinterland (except back thrusts). Also they should not have zigzag
trajectories, except for out-of-sequence thrusts. The line length and/or area should be
equal between deformed and restored; or if they are not then there should be proper
explanation.
We should restore one fault at a time, starting from the youngest fault
(Mukhopadhyay and Mishra 1999, 2005). At each stage of restoration, the restored
section must be admissible. If a section does not lead to an admissible restored section,
then we would need to suitably change the deformed-state section until an acceptable
restoration is obtained. The restoration process itself can potentially locate mistakes
and problem areas that may need modification.

165
References
Suggested reading

The following textbooks have influenced my thinking on structural geology over the
years. They always have pride of place on my desk. One should not be surprised if
treatment of any of the topics here is similar to that in any of these books.

Davis, G. H. (1984). Structural geology of rocks and regions. John Wiley & Sons, New
York.
Hobbs, B. E., Means, W. D. and Williams, P. F. (1976). An outline of structural
geology. John Wiley & Sons, New York.
Means, W. D. (1976). Stress and strain. Springer-Verlag, New York.
Ramsay, J. G. (1967). Folding and fracturing of rocks. McGraw-Hill, New York.
Ramsay, J. G. and Huber, M. I. (1983). The techniques of modern structural geology.
Vol. 1: Strain analysis. Academic Press, London.
Ramsay, J. G. and Huber, M. I. (1987). The techniques of modern structural geology.
Vol. 2: Folds and fractures. Academic Press, London.
Suppe, J. (1985). Principles of structural geology. Printice-Hall, New jersey.
Twiss, R. J. and Moores, E. M. (1992). Structural geology. W. H. Freeman and Co.,
New York

References cited

Anderson, E. M. (1951). The dynamics of faulting. Oliver and Boyd, Edinburgh, 206 pp.
Baby, P., Moretti, I., Guillier, B., Limachi, R., Mendez, E., Oller, J. and Specht, M.
(1995). Petroleum system of the northern and central Bolivian sub-Andean zone. In:
Petroleum Basins of South America (A. J. Tankard, R. Suarez S. and H. J. Welsink,
eds.), American Association of Petroleum Geologists Memoir., 62, 445-458.
Biswas, S. K. (1994). Status of exploration for hydrocarbons in Siwalik basin of India
and future trends. Him. Geol., 15, 283-300.
Boyer, S .E. and Elliot, D. (1982). Thrust Systems. American Association of Petroleum
Geologists Bulletin, 66, 1196-1230.
Chester, J. S. and Chester, F. M. (1990). Fault-propagation folds above thrusts with
constant dip. Journal of Structural Geology, 12, 903-910.
Cloos, E. (1955). Experimental analysis of fracture patterns. Geological Society of
America Bulletin, 66: 241-256.
Cloos, E. (1968). Experimental analysis of Gulf Coast fracture patterns. American
Association of Petroleum Geologists Bulletin, 52: 420-444.
Dahlstrom, C. D. A. (1969). Balanced cross sections. Canadian Journal of Earth
sciences, 6:743-757.

166
Davis, G. A., Anderson, J. L., Frost, E. G. and Shackelford, T. J. (1980). Mylonitization
and detachment faulting in the Whipple-Buckskin-Rawhide Mountains terrane,
southeastern California and western Arizona. Geological Society of America
Memoir, 153:79-129.
Effimol, I. and Pinezich, A. R. (1986). Tertiary structural development of selected
basins: Basin and Range province, northeastern Nevada. Geological Society of
America Special Paper, 208.
Elliott D., (1976). The energy balance and deformation mechanisms of thrust sheets.
Phil. Transactions of the Royal Society of London, 283A: 289-312.
Elliot, D. (1983). The construction of balanced cross-sections. Journal of Structural
Geology, 5: 101.
Fleuty, M. J. (1964). The description of folds. Proceedings of the Geological
Association of London, 75: 461-492.
Gansser, A. (1981). The geodynamic history of the Himalaya. In: Zagros-Hindukush-
Himalaya: Geodynamic evolution (H. K. Gupta and F. M. Delany, eds.), American
Geophysical Union, Washington, Geodynamic Series, 3: 111-121.
Gibbs, A. D. (1984) Structural evolution of extensional basin margins. Journal of the
Geological Society, 141: 609-620.
Hamblin, W. K. (1965). Origin of ‘reverse drag’ on the down-thrown side of normal
faults. Geological Society of America Bulletin, 76:1145-1164.
Jamison, W. R. (1987). Geometric analysis of fold development in overthrrust terranes.
Journal of Structural Geology, 9: 207-219.
Khalil, S. M. and McMclay, K. R. (2002). Extensional fault-related folding,
northwestern Red Sea, Egypt. Journal of Structural Geology, 24: 743-762.
Mandl, G. (1984). Rotating normal faults – the bookshelf mechanism. American
Association of Petroleum Geologists Bulletin, 68: 502-503.
Mandl, G. (1987). Tectonics deformation by rotating parallel faults: the “bookshelf”
mechanism. Tectonophysics, 141: 277-316.
Marshak, S. and Woodward, N. (1988). Introduction to cross section balancing. In:
Basic methods of structural geology (S. Marshak and G. Mitra, eds.), Prentice Hall,
Englewood Cliffs, New Jersey, 303-332.
McClay, K. R. (1992). Glossary of thrust tectonics terms. In: Thrust tectonics (K. R.
McClay, ed.), Chapman and Hall, 419-433.
McNaught, M. A. and Mitra, G. (1993). A kinematic model for the origin of footwall
synclines. Journal of Structural Geology, 15: 805-808.
Medwedeff, D. A. and Suppe, J. (1997). Multibend fault-bend folding. Journal of
Structural Geology, 19: 279-292.
Mishra, P. and Mukhopadhyay, D. K. (2002). Balanced structural models of Mohand
and Santaugarh ramp anticlines, Himalayan foreland fold-thrust belt, Dehra Dun
recess, Uttaranchal. Journal Geological Society of India, 60: 649-661.
Mitra, S. (1986). Duplex structures and imbricate thrust systems; Geometry structural
position and hydrocarbon potential. American Association of Petroleum Geologists
Bulletin, 70: 1087-1112.

167
Mitra, S. (1990). Fault-propagation folds: geometry, kinematic evolution, and
hydrocarbon traps. American Association of Petroleum Geologists Bulletin, 74: 921-
945.
Mitra, S. 1992. Balanced structural interpretations in fold and thrust belts. In: Structural
Geology of Fold and Thrust Belts (S. Mitra and G. W. Fisher, eds.), The John
Hopkins University Press, Baltimore, 53-77.
Mitra, S. and Namson, J. S. (1989). Equal-area balancing. American Journal Science.,
289, 563-599.
Morley, C. K. (1988). Out-of-sequence thrusts. Tectonics, 7: 539-561.
Mukhopadhyay, D. K. and Mishra, P. (1999). A balanced cross section across the
Himalayan foreland belt, the Punjab and Himachal foothills: A reinterpretation of
structural styles and evolution. Proceedings Indian Academy of Sciences (Earth
Planetary Sciiences), 108: 189-205.
Mukhopadhyay, D. K. and Mishra, P. (2005). A balanced cross section across the
Himalayan frontal fold-thrust belt, Subathu area, Himachal Pradesh, India; thrust
sequence, structural evolution and shortening. Journal of Asian Earth Sciences, 25:
735-646.
Passchier, C. W. and Simpson, C. (1986). Porphyroclast systems as kinematic
indicators. Journal of Structural Geology, 8: 831-843.
Peacock, D. C. P., Knipe, R. J. and Sanderson, D. J. (2000). Glossary of normal faults.
Journal of Structural Geology, 22: 291-305.
Platker (1965). Tectonic deformation associated with the 1964 Alaska earthquake.
Science, 148: 1685.
Price, R. A. (1981). The Cordilleran foreland thrust and fold belt in the southern
Canadian Rocky Mountains. Geological Society of London Special Publication, 9:
427-448.
Ramsay, J. G. (1980). Shear zone geometry: a review. Journal of Structural Geology, 2:
83-89.
Suppe, J. 1983. Geometry and kinematics of fault-bend folding. American Journal
Science, 283: 684-721.
Suppe, J. and Medwedeff, D. A. 1984. Fault-propagation folding. Geological Society of
America Abstract with Program, 16: 670.
Suppe, J. and Medwedeff, D. A. 1990. Geometry and kinematics of fault-propagation
folding. Eclog. Geol. Helv., 83, 409-454.
Walsh, J.J. and Watterson, J. (1988). Dips of normal faults in British coal measures and
other sedimentary sequences. Journal of the Geological Society, 145: 859-874.
Wise, D. U., Dunn, D. E., Engelder, J. T., Geiser, P. A., Hatcher, S. A., Kish, S. A.,
Odom, A. L. and Schamel, S. (1984). Fault-related rocks: suggestions for
terminology. Geology, 12: 391-394.
Woodward, N. B., Boyer, S. E. and Suppe, J. (1989). Balanced geological cross-
sections: An essential technique in geological research and exploration. American
Geophysical Union Short Course, 6:1-132.
Xiao, H and Suppe, J. (1992). Origin of rollover. American Association of Petroleum
Geologists Bulletin, 76:309-529.

168

Das könnte Ihnen auch gefallen