Structural Geology

Lecture Notes
Professor Andrew Hynes

Structural Geology
Lecture Notes

Professor Andrew Hynes

CONTENTS
EVALUATION AND GRADING ..............................................................................................................................1
GENERAL TEXTS .....................................................................................................................................................1
GENERAL LABORATORY INSTRUCTIONS ......................................................................................................3
DESCRIPTIVE STRUCTURAL GEOLOGY ..........................................................................................................4
ATTITUDES OF PLANES .......................................................................................................................................4
ATTITUDES OF LINES...........................................................................................................................................5
APPARENT DIP.......................................................................................................................................................5
RELATIONSHIP OF APPARENT TO TRUE DIP ..................................................................................................6
TRUE DIP AND STRIKE FROM TWO APPARENT DIPS ....................................................................................8
THREE-POINT PROBLEMS .................................................................................................................................10
STRATIGRAPHIC THICKNESS ..........................................................................................................................11
STEREOGRAPHIC PROJECTIONS .....................................................................................................................14
PLANES AND TOPOGRAPHY ..............................................................................................................................21
SEDIMENTARY STRUCTURES ...........................................................................................................................23
BEDDING-PLANE STRUCTURES ......................................................................................................................24
UNCONFORMITIES .............................................................................................................................................25
SOFT-SEDIMENT FOLDS ....................................................................................................................................26
DIAPIRISM ...............................................................................................................................................................28
SEDIMENTARY FABRIC .......................................................................................................................................28
PRIMARY IGNEOUS FEATURES ........................................................................................................................30
PYROCLASTIC ROCKS .......................................................................................................................................32
INTRUSIVE BODIES ............................................................................................................................................33
TECTONIC FEATURES..........................................................................................................................................35
FAULTS .....................................................................................................................................................................36
SOLUTION OF FAULT PROBLEMS ....................................................................................................................39
FAULT SOLUTION ON MAPS .............................................................................................................................42
STRESS ......................................................................................................................................................................44
MOHR STRESS CIRCLE.......................................................................................................................................46
NAVIER-COULOMB FAILURE CRITERION.....................................................................................................47
STRAIN ......................................................................................................................................................................49
DETERMINATION OF STRAIN ..........................................................................................................................53
PROGRESSIVE STRAIN ......................................................................................................................................58
BEHAVIOUR OF MATERIALS .............................................................................................................................64

ii
EFFECT OF PRESSURE AND TEMPERATURE.................................................................................................65
TIME-DEPENDENT STRAIN ...............................................................................................................................65
FOLDS AND FOLDING ..........................................................................................................................................68
(A) FOLDING OF A SINGLE SURFACE ........................................................................................................................68
(B) MANY SURFACES ...............................................................................................................................................69
(C) ATTITUDES OF FOLDS .........................................................................................................................................69
(D) FOLD STYLE .......................................................................................................................................................70
FOLD PROJECTION ...............................................................................................................................................75
CLEAVAGE ..............................................................................................................................................................82
LINEATIONS ............................................................................................................................................................85
CLEAVAGE AND FOLDING .................................................................................................................................86
FOLDS ON THE STEREOGRAM..........................................................................................................................89
FOLD MECHANISMS .............................................................................................................................................92
JOINTING .................................................................................................................................................................96
CLEAVAGE & LINEATION ON A STEREOGRAM ..........................................................................................99
CONTOURING ON A STEREOGRAM ..............................................................................................................100
DRILL HOLE PROBLEMS...................................................................................................................................101
POLYDEFORMATION .........................................................................................................................................105
POTENTIAL FOR CONICAL FOLDS ................................................................................................................108
RELATIVE AGES OF FABRIC ELEMENTS .....................................................................................................111
INTERFERENCE PATTERNS IN PLANAR SECTION .....................................................................................113
SHEAR-ZONE DEFORMATION .........................................................................................................................117
DEFORMATION MECHANISMS AND MICROFABRIC ...............................................................................119

EPSC 203 (2012) p. 1

COURSE NOTES

NOTICES TO STUDENTS
In accord with McGill Universitys Charter of Students Rights, students in this course have the right to
submit in English or in French any written work that is to be graded.
McGill University values academic integrity. Therefore all students must understand the meaning and
consequences of cheating, plagiarism and other academic offences under the Code of Student Conduct
and Disciplinary Procedures (see www.mcgill.ca/students/srr/honest/ ) for more information).

EVALUATION AND GRADING

The grading system is as follows:
Laboratories

Theory

Mid-term test
(3-hour test in place of lab.)
Final examination
(last lab session of term)
Final examination
(3-hour formal exam)

20%
40%

40%
_____
100%

GENERAL TEXTS
There are many general texts in introductory structural geology. None is suitable for
everyone. All those listed below are available in the library. I do not recommend a text for the
course. This is primarily because, although many of the texts are very good, there is no one
text that usefully combines the theoretical aspects of the subject with the laboratory aspects.
For laboratory work, the best text is that by Marshak and Mitra. Ragan is an alternative.
Books by Davis, Suppe, van der Pluijm and Marshak and Twiss and Moores provide the best
theoretical treatments. In general, you should be able to follow this course using these course
notes rather than a text, although you may wish to consult one or more of these books to
supplement them.
Billings MP 1972 Structural geology. Prentice-Hall Inc., Englewood Cliffs, N.J. 606 p.
The classic elementary text in structural geology. Somewhat elementary, and
long-winded in places, but very useful for students who are having problems.

EPSC 203 (2012) p. 2

Davis GH 1996 Structural geology of rocks and regions. John Wiley, 776 p.
Well illustrated and up-to-date. Includes a section on plate tectonics.
Dennis JG 1972 Structural geology. Ronald Press Co., New York, 532p.
Good all-round text, especially for terminology. Very concise.
Hobbs BE, Means WD, and Williams PF 1976 An outline of structural geology. Wiley, 571p.
This text is particularly strong in its treatment of microfabric, and contains a useful
description of a variety of tectonic associations.
Marshak S and Mitra G 1988 Basic methods of structural geology. Prentice Hall, 446p.
van der Pluijm, BA and Marshak, S 2004 Earth Structure; An introduction to structural geology
and tectonics, 2nd edition. Norton & Co., New York, 656p.
Comprehensive text treating both basic structural geology and regional tectonics.
Ragan DM 1985 Structural geology: an introduction to geometrical techniques. 3rd Edition.
Wiley, 393 p.
Detailed discussion of laboratory techniques.
Ramsay JG 1968 Folding and fracturing of rocks. McGraw-Hill, 568 p.
More advanced general text. Mathematical treatment of stress and strain.
treatment of folds and cleavage in the later chapters.

Good

Spencer EW 1977 Introduction to the structure of the earth. McGraw-Hill, 640 p.

Good all-round text.
Suppe J 1985 Principles of structural geology. Prentice-Hall, 537 p.
Well-illustrated introductory text. Includes useful chapters summarising the geology of
the Appalachians and the Cordillera.
Twiss RJ & Moores EM 1992 Structural Geology. W. H. Freeman, NY, 532 p.
Turner FJ and Weiss LE 1963 Structural analysis of metamorphic tectonites. McGraw-Hill,
545 p.
Advanced general text. Although sometimes obscurely written, it remains one of the
best descriptions of practical structural analysis.

EPSC 203 (2012) p. 3

GENERAL LABORATORY INSTRUCTIONS
(a) Equipment
Students should come to the first, and all subsequent, laboratories equipped with the
following:
drafting compass
semicircular (or circular) protractor
ruler or scale
triangle (approx. 15x7 cm)
calculator with trig functions
pencils: 2H and HB
eraser
colouring pencils
plain paper
tracing paper (or at least translucent paper)
(b) Nature and complexity of exercises
The laboratory exercises are designed to complement the lectures. The subjects of
the exercises will have been treated in lectures before they are introduced in the laboratory.
In certain instances, sample problems will have been solved in class beforehand. In all other
cases the student should be able to devise a suitable method for the solution alone, or with
help from one of the demonstrators. Students are encouraged to seek the assistance of the
demonstrators during the laboratory hours.
(c) Completion and marking
You are expected to work on the exercises during the laboratory and, if necessary, to
complete the problems during your own time. You may submit completed solutions to be
corrected by the demonstrators, but for this they must be received by the end of the
subsequent laboratory session. These solutions will be corrected, but not graded. They do
not count towards your final mark in the course. The responsibility for gaining facility with the
lab material thus rests squarely with you. You are strongly encouraged, however, to make
use of the demonstrators to monitor your progress.
(d) Website
Lab exercises for each week will be available at the first lecture each week. They will
also be posted on the website (www.eps.mcgill.ca/~courses/c203) on the Monday preceding
the lab session. Solutions to the lab will be available at the subsequent lab session and will
also be posted on the web.

EPSC 203 (2012) p. 4

DESCRIPTIVE STRUCTURAL GEOLOGY
The first stage of any structural analysis is a description of the structures observed.
Since all 3D structures may be considered to be arrangements of planes and lines in space
this involves the specification of the attitudes and positions of planes and lines.
ATTITUDES OF PLANES
The attitude of a plane is commonly expressed by a "strike" and a "dip".
(a) Strike = orientation (bearing) of a horizontal line in the plane.
Orientation is commonly expressed as a number between 0o and 360o, where the
number of degrees is measured clockwise, beginning at north (thus east is 090o, south is 180o
etc.) It may also, however, be expressed as a "bearing", e.g. N20oE (= 020o), N20oW (=
340o). It is always a good idea to express a bearing as a 3-digit number (e.g., 020 rather than
20) to avoid possible confusion of bearings of less than 100 with angles measured in the
vertical plane, such as dips (see below).
Strike has no direction, only orientation. i.e., a plane striking E may also be described
as striking W. A strike of 060o is exactly equivalent to a strike of 240o, and so on.
(b) Dip line = line perpendicular to strike, lying in the plane.
(c) Dip = angle between dip line and its horizontal projection.

30

ik
str

ine
e-l

0
03

horiz. projn. of dip

dip ( 40 )
dip
-lin
e
o

0
21

Figure 1

To specify the attitude of a plane completely it is necessary to specify its strike and its
dip and, because strike has no direction, the (approximate) direction of the dip line (always
down-dip line) must also be given.

EPSC 203 (2012) p. 5

e.g., plane shown on Fig. 1 (solid lines) strikes 030o, dips 40oSE or strikes 210o, dips
40oSE.
(d) Right-hand rule
To avoid having to specify the direction of dip it is commonly assumed that the strike
has a direction such that when the plane is viewed in that direction it is dipping off to the right.
According to this "right-hand" rule the plane depicted on Fig. 1 would have a strike of 030o
not 210o, and the dip would be 40o.
(e) Dip/dip bearing
Another unambiguous and brief shorthand is to describe the plane by means of its dip
and "dip bearing", where the dip bearing is the orientation of the horizontal projection of the
dip-line. The dip bearing has a direction down the dip. Thus, the plane of Fig. 1 has a dip of
40o and a dip bearing of 120o.
(f) Pole to a plane
The attitude of a plane may also be specified uniquely by describing the attitude of its
"pole"; the line that is normal to it. This is in fact the most concise way in which to describe a
plane, and it is the most commonly used method in advanced structural work.
ATTITUDES OF LINES
The attitudes of lines are described by their "trend" and "plunge".

(a) Trend = orientation of horizontal

projection of line, with a direction which is
the direction in which the line goes
downwards. (The dip bearing is the trend of
the dip line)
(b) Plunge = angle between line and its
horizontal projection, measured downwards
from the horizontal. (The dip is the plunge of
the dip line; Fig. 2)

Figure 2

APPARENT DIP
If a plane is cut in any fashion it will appear on the cut surface as a line, unless the cut

EPSC 203 (2012) p. 6

is precisely parallel to the plane. This line is the "trace" of the plane on the cut surface. Its
plunge is an "apparent dip" for the plane; i.e., the dip of the plane as it appears to be on the
cut surface (Fig. 3).
The apparent
dip cannot be larger
than the true dip of
the plane, and it is
the same as the true
dip only if the plane
has been intersected
along its dip line.
(Another
way
of
stating this is to say
that the lines with the
steepest plunge in a
plane are all dip
lines, and they have
a plunge that equals
the true dip. All other
lines have shallower
plunge; the limiting
case being the strike
lines, whose plunge
is zero.)

Figure 3 Plane (solid lines) showing apparent dips in two

different directions. is the true dip of the plane. 1 and 2 are
apparent dips in two different directions.

PITCH OF LINE (IN A PLANE)

It is sometimes convenient to
specify the attitudes of lines, occurring in
planes, in terms of their "pitch", which is the
angle between the line and the strike line of
the plane, measured in the plane. The
attitude of a line cannot be determined from
its pitch unless the attitude of the plane in
which the pitch was measured is also
known (Fig. 4).
Figure 4
RELATIONSHIP OF APPARENT TO TRUE DIP
Let be the true dip, be the apparent dip (Fig. 5).
(a) Trigonometric relationship

EPSC 203 (2012) p. 7

ABCD is the plane of interest. ABC'D' is horizontal. The apparent dip in any direction AC,
at angle to the strike line is easily determined:

tan =

h h l
= . = tan sin ,
d l d

where = 0, tan =0; = 0.

Figure 5

(b) Orthographic Relationship

The problem above may also be solved by an orthographic method. An orthographic
projection is a projection in which straight lines project as straight lines and angular
relationships are preserved. Using these types of projection, many of the planes in Fig. 5 are
represented on the horizontal plane.
This is done by rotating each such plane about the horizontal line in it, until the plane is
horizontal.
PROCEDURE (assuming is known. See Fig. 6)

EPSC 203 (2012) p. 8

1. Construct ABC'D'. Since this is already horizontal it is a simple rectangle.
2. Construct triangle BC'C, knowing , and that angle BC'C is 90o. (Call this C C1). C'C = h in
the figure.
3. Knowing h, we may now construct triangle
AC'C by drawing AC', constructing a line C'E
perpendicular to AC', and marking off a
distance h on it, to give C (=C2). Angle
C'AC2 = .
The method may easily be reversed
to determine the true dip, if the strike and
one apparent dip are known.

C1

C
E

Figure 6
TRUE DIP AND STRIKE FROM TWO APPARENT DIPS
Consider the plane depicted on Fig. 7 by the solid lines ABCDE. 1, 2 and (on the
horizontal plane) are known. and are unknowns.

(a) Trigonometric method

tan 1 =

h
h
; tan 2 = .
d
n

From AE'D':

l l d
cos [ - (90 - )] = = . .
n d n
Therefore

cos [ - (90 - )] =

l d h sin tan 2
. . =
d h n
tan 1

EPSC 203 (2012) p. 9

cos [-(90-)] = cos cos (90-) + sin sin (90-)
= cos sin + sin cos
Therefore
But

cos +

sin tan 2
=
tan tan 1

which may be solved for .

Then

h h d tan 1
tan = = . =
l d l
sin
from which may be found.

B
C
A
D

D
E

Figure 7

(b) Orthographic Method (Fig. 8)

1. Construct AC' and AE', o apart.
2. Construct the orthographic projection of triangle E'E, by laying off AE' at 2 to AE' and
dropping a perpendicular to AE' at E' with EE'=h (as in figure).

EPSC 203 (2012) p. 10

Figure 8

3. Do the same for triangle ACC'. Note that the position of C' is determined by the fact that
EE'=CC'=h. C'E' is the strike line for the plane.
4. Construct triangle AD'D, where angle AD'D = 90o and D'D=h. Angle D'AD is .

THREE-POINT PROBLEMS
A "three-point problem" is a problem in which the positions of three points on a plane
are known, and the attitude of the plane must be determined. A three-point problem may be
reduced to a problem in which two apparent dips for the plane are known, which may be
solved as above.
As an example, consider the problem in which the three points A, B and C are known
to occur on the same surface (e.g., the top of a layer of limestone). All three points are in fact
beneath the present land surface and their positions were determined through drilling. In Fig.
9, the positions directly above B and C on the horizontal surface at the elevation of A are
marked B and C. Triangles ABB and ACC are orthographic projections of vertical triangles
beneath AB and AC.
Line AB on the map is horizontal (as are all lines on a map), but the true line AB, of
which this is the horizontal projection, plunges at an angle which may be determined from the
difference in elevation between A and B and the length of the horizontal projection of AB
(measured on the map). These two lengths allow the construction of a right-angled triangle
from which the plunge of AB may be determined, either trigonometrically or orthographically
(see Fig. 9, where B is the projection of B, so that AB is the projection of AB into the
horizontal surface at A. AB is the true slope distance). A similar procedure may be used to
determine the plunge of AC. We then have two apparent dips.

EPSC 203 (2012) p. 11

A more direct way of solving the
problem graphically would be to
determine the distance along the
horizontal projection AC for which the
surface had dropped to the elevation of
B. Since the drop from A to B is 200 m,
and that from A to C is 300 m, this
distance is 2/3 of AC (point P, Fig. 10).
Point P is now directly above a point on
the plane that is at the same elevation
as B. BP is therefore a strike line for
the plane, and may be used to
determine the strike. The dip may then
be
determined
by
orthographic
methods, bearing in mind that the dip
bearing must be perpendicular to the strike (Fig. 10).

Figure 9

Figure 10

STRATIGRAPHIC THICKNESS
A stratum is a layer of rock of uniform character. Strata are generally bounded by
approximately plane surfaces parallel to each other. The thickness of the stratum, or
"stratigraphic thickness" is the thickness of the layer measured perpendicular to the
bounding surface. Any transect of the stratum in a direction other than parallel to the normal
to the surface gives an apparent thickness that is greater than the true thickness. It must
therefore be adjusted to the true thickness. This adjustment may be done by either graphical
or trigonometric methods.

EPSC 203 (2012) p. 12

The normal to a surface has the property that it is perpendicular to the trace of the
surface in any plane that contains the normal. The perpendicular to the trace of a surface in
an arbitrarily chosen plane (such as an outcrop surface or map projection) is not, however, in
general the normal to that surface. The true normal to the surface may be determined from
this perpendicular by construction of another section that contains it, and projecting the
perpendicular into the position in which it is again perpendicular to the (new) trace of the
surface in the (new) section. Determination of the normal to a surface therefore in general
involves two projections of the apparent thickness.

Figure 11

As an example, consider a stratum of limestone that strikes at 030o and dips 50o to the
SE. This limestone is transected in a traverse AB oriented due east, and has an apparent
thickness of 120 m on the traverse (Fig. 11).
In map view, the apparent thickness AB may be adjusted to the perpendicular
thickness BP, either graphically or from:
BP = AB sin 60
BP is not, however, the normal to the surface of the limestone. A vertical section
containing BP reveals that BP is at 50o to the trace of the surface (Fig. 12).

EPSC 203 (2012) p. 13

This apparent thickness BP may be adjusted to the true thickness BQ, again
graphically, or from:
BQ = BP sin 50

The true thickness is therefore 120 sin 60 sin 50 = 80 m

Figure 12

EPSC 203 (2012) p. 14

STEREOGRAPHIC PROJECTIONS
Stereographic projection provides a convenient means by which to represent the
attitudes of lines and planes in space. It is a powerful technique, and is widely used in
structural geology. The only major disadvantage of the technique is that only attitude may be
represented on a stereographic projection; position cannot be shown. Two lines with the
same attitude plot at the same place on the projection, regardless of their positions in space.
As a result, it is commonly necessary to use stereographic projections in conjunction with
scaled two-dimensional sections or with sketches from which trigonometric relationships may
be extracted. Stereographic projections, however, provide by far the quickest means by
which to determine angular relationships within the desired two-dimensional section.
In
the
stereographic
projection,
all
attitudes are
considered with reference to a
sphere, the top of which
represents the up-direction. The
equatorial plane of the sphere is
divided into 360o with N, S, E and
W in the usual way. The line or
plane which is to be projected is
considered to pass through the
centre of the sphere (it is because
of this that its position is no longer
characteristic). The position and
orientation of the line or plane
within the sphere are now fully
specified.
The stereographic
projection
reduces
all
the
three-dimensional data into a
two-dimensional
diagram
or
"stereogram" in which all lines plot
as points, and all planes plot as
parts of circles. The stereogram is
prepared by projecting every
feature onto the equatorial plane
of the sphere in a two-stage
procedure:

Figure 13 AB: strike line of plane; APB: Intersection

of plane with lower hemisphere; AQB: projection of
1. The intersection of the feature plane onto equatorial plane; P: where line hits lower
with the lower hemisphere of the hemisphere; Q: projection of P into equatorial plane.
sphere is determined. In the case
of a line this is a point. In the case of a plane this is half a great circle.
2. The lower hemisphere intersection is projected back towards the point at the top of the

EPSC 203 (2012) p. 15

sphere until it intersects the equatorial plane.
Note that the projection depicted on Fig. 13 is a "lower hemisphere" projection. This is
the kind commonly used in structural geology. It is equally possible to project the feature
upwards to the upper hemisphere, and then project it downwards toward the lower point of
the sphere. This "upper hemisphere" projection is the one commonly used in crystallography.

On the stereogram that results from

this projection technique, lines plot as points.
The radius on which the line lies is
determined by the trend of the line, and the
position on the radius is determined by its
plunge. Lines with shallow plunge plot close
to the perimeter of the stereogram (known
as the "primitive"); lines with steep plunge
plot close to the centre. In Figure 14, Q is a
line plunging P at 120, cf. Figure 13.

Figure 14

EPSC 203 (2012) p. 16

Figure 15

Planes plot as circular arcs that

intersect the primitive on diameters of the
sphere. The orientation of the diameter is
determined by the strike of the plane. The
curvature of the arc is determined by its dip:
shallowly-dipping planes have large
curvature (small radius of curvature)
reflecting the fact that all lines that lie in
them have shallow plunge, the limiting case
being that of a horizontal plane for which the
arc coincides with the primitive. Steeply
dipping planes have small curvature, the
limiting case being that of a vertical plane for
which the arc has no curvature and
coincides with the diameter. These circular
arcs are the projections of great circles on
the reference sphere, and are referred to as
"great circles".

Stereograms that are calibrated at 1o or 2o intervals, known as "stereonets", are

available commercially, so that lines and planes may be plotted rapidly on overlays. These
stereonets are of two kinds. The first, known as the "true stereographic projection" or "Wulff
net", is prepared on exactly the principles described above. It has the advantage that all lines
that are equal angular distances from a given line (i.e., that lie on a cone) plot as points on a
perfect circle. It has, however, the disadvantage that angular distances appear much greater
near the primitive than near the centre of the stereogram (e.g., two points representing lines
5o apart are much more widely separated if they have shallow plunge than if they have steep
plunge). A net that has been distorted to compensate for this (rarefied near its centre and
condensed near its primitive) so that angular distances are the same throughout the net is
known as a "Lambert", "Schmidt" or "equal area" net. On this net, a set of lines lying on a
cone appears as a distorted circle. The equal area net is the one most commonly used in
structural geology, although for most applications including all those discussed in the next
few pages either could be used.
PLOTTING ON THE STEREOGRAM
Stereograms are commonly prepared on translucent paper overlays on stereonets.
Since it is necessary to rotate the overlay on the stereonet it is attached to the stereonet by a
pin passing through the centres of both. The intersection of the "north" radius with the
primitive is marked on the overlay. If you are in the habit of using both Wulff and equal-area
nets it is advisable also to identify which type of net is being used.
LINES

EPSC 203 (2012) p. 17

The stereogram may be considered
to have 360 radii, each of which represents
a vertical section with a strike corresponding
to the azimuth of the radius. A line must
necessarily lie in the vertical section whose
strike is the same as its trend. To identify
this section on the overlay, orient the overlay
with its north radius coincident with that of
the stereonet and mark on the overlay the
position of the radius whose azimuth is that
of the strike of the line (as seen on the
stereonet underneath the overlay).
To
determine where on this vertical section the
line should plot it is necessary to calibrate
the vertical section for plunge. This may be
done by rotating the overlay until the section
o
is superimposed on any of the east-west or Figure 16 Projection of a line plunging 30 at
north-south radii on the stereonet, because 050. The radius at 050 corresponds to the
these radii are marked on the stereonet, and vertical section trending at 050.
are divided into degrees.
The correct
position of the line on the section (a point) may now be determined by counting in along the
section from the primitive the number of degrees corresponding to the plunge of the line (Fig.
16). The overlay may now be restored to its original position, and the stereogram is complete.
PLANES
A plane appears on the stereogram
as part of a circle, intersecting the primitive
at two points on the same diameter. This
diameter has the same azimuth as the strike
of the plane in question, and may be found
in a manner analogous to that for the trend
of a line. On the stereonet, the possible
forms of great circles corresponding to
planes are drawn for a strike of 000o. In
order to draw the great circle for the plane in
question, therefore, the overlay must be
rotated until the strike of the plane is
oriented north on the stereonet.
The
appropriate great circle may then be
selected, by counting in from the primitive
the number of degrees corresponding to the
dip of the plane. The great circle may be
drawn, and the overlay restored to its
Figure 17 Plane striking 050, dipping 60oNW.
original position (Fig. 17).

EPSC 203 (2012) p. 18

A plane, represented by a line on the
stereogram, contains an infinite number of
lines, represented by points on the
stereogram. Each point on the great circle
that
represents
a
plane
therefore
corresponds to a line within the plane. A
point on the great circle that is also on the
primitive has zero plunge, and therefore
represents the strike line for the plane. Note
that the strike line might be represented by
either of the two such points (at the two
ends of the diameter), reflecting the fact that
the strike has orientation but no direction.
Points on the great circle that do not lie on
the primitive have finite plunge.
The
magnitude of the plunge increases as
distance from the primitive increases,
reaching a maximum at the point
Figure 18 Plane striking 120, dipping 65oSW.
representing a line with trend 90o from the
strike of the plane. This point represents the
dip line for the plane. The angle between the strike-line and the dip line must be measured in
the plane common to both of them, which is the plane itself. Angles in this plane may be
measured by once again rotating the overlay until the great circle is superimposed on one of
those drawn on the stereonet, at which position the great circle is calibrated for degrees. The
angle between the dip line and the strike line is 90o as expected (Fig. 18).
APPARENT DIP
As noted above, any line that lies on a plane appears as a point on the great circle
corresponding to the plane on a stereographic projection. The apparent dip of the line is its
plunge. Determination of the apparent dip of a line with a given trend in a plane therefore
reduces to the problem of locating the point corresponding to that line in the projection of the
plane, and determining its plunge.
e.g., Consider a plane striking 025o and dipping 50o to the ESE. We wish to determine the
apparent dip of that plane in a vertical section striking at 060o.

EPSC 203 (2012) p. 19

1. Plot the projection of the plane (Fig. 19).

2. Plot the vertical section. The apparent dip
is the line that is common to both planes.
3. Determine the plunge of the line, by
rotating the overlay until the line lies on one
of the calibrated vertical sections of the
stereonet and counting in the plunge from
the primitive. (Apparent dip is 34o)

Figure 19
PLANE FROM TWO APPARENT DIPS
As we have seen, the attitude of a plane is uniquely determined by the attitudes of two
lines that lie within it (e.g., by two apparent dips). On the stereogram, the attitude of the plane
may be determined by plotting the projections of the two lines (as points) and finding the great
circle that contains both points. This is then the projection of the plane.

EPSC 203 (2012) p. 20

e.g., A plane has an apparent dip of 50oNE
in a vertical section striking at 040o, and an
apparent dip of 30oS in a vertical section
striking at 170o. Determine its attitude.
1. Plot both vertical sections, and mark on
them the points corresponding to the two
apparent dips (Fig. 20).
2. Rotate the overlay until both points fall on
the same great circle on the stereonet, and
mark that great circle on the overlay.
3.
Read off the dip of the plane
corresponding to the great circle.
4. Rotate the overlay back to the reference
position and read off the strike of the great
circle. (Answer : strike 006o, dip 65oE)

Figure 20

PITCH
The pitch of a line in a plane is
measured in the plane itself. Since the
great circles on the stereonet are calibrated
in degrees, it is possible to determine the
angular relationships within a plane by
rotating the overlay until the plane overlies a
great circle on the stereonet and reading off
the angle in that great circle. The pitch of a
line in a plane is the angle between the line
and the strike of the plane.
e.g., (Fig. 21) A line trending at 220o in a
plane striking 010o and dipping 50oW
plunges 32o and pitches 43oS.

Figure 21

EPSC 203 (2012) p. 21

PLANES AND TOPOGRAPHY
Any flat plane intersects any other flat plane in a straight line, the attitude of which may
be determined from a stereogram. The line that marks the intersection is known as a "trace"
(of the second plane in the first or the first plane in the second). On a map, plane layering will
appear as straight lines - the traces of the layering on the horizontal plane - provided there is
no topographic relief. In this case the traces are the strikes of the layering. If there is
topographic relief, however, even a completely planar set of layering will have traces on a
map that are not straight. These traces are lines joining points at which the erosion surface
and the layer are at the same elevation.

For horizontal layering these traces

are parallel to the topographic contours, so
that the contacts of the layers follow the
contours exactly (Fig. 22).
For vertical layering the traces are
straight lines parallel to the strike of the
layering; they are unaffected by topography
(Fig. 23).
For layering with dips other than 0o or
Figure 22 Horizontal bedding.
90 the relationship between topography and
Limestone/sandstone contact at 400 m;
the traces is more complex. It is most simply
sandstone/shale contact at 250 m.
analyzed by the use of strike lines - lines
parallel to the strike of the layering - spaced
at equal intervals reflecting elevations of the layer that are the same as those of the contour
intervals. Strike lines drawn in this way are structure contours on the surface of the layer.
They are closely spaced when the dip is steep and widely spaced when it is shallow. Their
spacing may be determined by a simple trigonometric calculation if the dip of the layering and
the scale of the map are known. Using the strike lines, points where the layer and the erosion
surface intersect may be identified, and joined to produce the trace of the layer on the map.
Alternatively, the trace of a layer on a map may be used to identify the bearing and spacing of
strike lines and hence the attitude of the layering (Fig. 24).
o

EPSC 203 (2012) p. 22

Figure 23 Vertical bedding

Figure 24 Strike lines on a sandstone/shale contact dipping 30o towards the SSW. In
this figure, the straight lines labelled '200', '300' and '400' are strike lines at those

EPSC 203 (2012) p. 23

SEDIMENTARY STRUCTURES
Sedimentary structures are those produced by the processes of sedimentation and
lithification, as distinct from those produced in sedimentary rocks by deformation after
consolidation. They provide the most widely used indicators of the way-up or "facing"
directions of sedimentary successions.
(a) stratification (bedding) = layering parallel to the surface of deposition. "Lamination" =
layering with thickness < 1 cm.
Stratification is caused by one or more of: seasonal changes, catastrophic effects
(storms, mud-slides etc.), current changes (directional, speed or both), source area
disturbances, compensation level changes etc.
Its structural significance is that it
approximates the paleo-horizontal, except in the case of cross-bedding.
(b) cross-stratification (cross-bedding) = two or more sets of stratification with different
attitudes.
Cross-bedding is produced by
deposition of material on prograding slopes
such as those of dunes and ripples, followed
by a change in flow regime that results in the
deposit of a new (horizontal) bed on top (Fig.
25a).
Change to the new flow regime is
commonly accompanied by some erosion of Figure 25 Formation of cross-bedding.
already deposited material. This results in
truncation of the cross-bed set on its upper
side, never on its lower side (Fig. 25b).
Cross-bedding is an extremely useful "facing" (younging) criterion. i.e. it indicates
which way up the beds are (after deformation).
(c) graded bedding = variation in grain size from bottom to top of bed.
"Normal" grading is that in which the coarsest grains are at the base. It is due to the
more rapid settling of coarse grains as a result of their relatively low surface-area/volume
ratio. Graded beds are commonly produced by a sudden influx of poorly sorted (variable
grain size) material into a depositional basin. They are generally recognised by the sharp
lithological boundary at their bases and by the colour contrast within them (usually pale at the
base and darker at the top). Their value as a facing criterion is self-evident (Fig. 26a).
"Reverse" grading is not a common feature, but it may be produced in a number of
environments:

EPSC 203 (2012) p. 24

(i) on beaches, due to the sieving

effect when beach-sands are
reworked by waves.
(ii) in fluviatile environments, as a
result of the gradual speeding up
of a sediment- charged current of
water. i.e., when travelling slowly
it drops everything; as it speeds up
it drops only the coarser fraction.
The result is that, although the
mean grain size increases up-bed,
the configuration is not exactly the
inverse
of
normal
grading,
Figure 26 a Normally graded beds. b Reversely graded because the "finer" part of the bed
contains some coarse clasts (Fig.
beds.
26b).

Grading should not be used as a facing criterion in sedimentary environments in which

reverse grading might be expected. In practice, this is rarely a problem, and grading is the
most valuable facing criterion, especially since it may be seen even after quite extensive
metamorphic recrystallization.
BEDDING-PLANE STRUCTURES
(a) Sole marks
Sole marks are structures observed in the sole (base) of a bed due to the filling of an
irregularity in the bed beneath. There are many different kinds, of which the most common
are:
(i) Flute casts: fillings (casts) of the depressions excavated in the surface of an
unconsolidated sediment by current eddies.
(ii) Scour marks (casts): fillings of grooves due to the dragging of a large fragment across
unconsolidated sediments.

EPSC 203 (2012) p. 25

(iii) Load casts: bulbous protrusions on the
underside of a bed, due to differential
sinking of the sediment in the bed into the
layer below (Fig. 27a).
All three of the above are most
commonly observed in the bases of graded
beds (or sand beds) where they overlie
much finer-grained mudstones.
(iv) Animal tracks.
Sole marks provide clear and
unambiguous facing criteria and, in some
cases, current directions.
(b) Sandstone balls.

Figure 27 Formation of load structures (a) and

sandstone balls (b).

Sandstone balls are isolated balls of

sandstone in an underlying mud. This is an extreme case of load casts (Fig. 27b).
(c) Ripples.
Ripples may be symmetric or asymmetric ripples, depending upon the current scheme
that generated them, but are difficult to use as facing criteria.
(d) Mudcracks.
Mudcracks are cracks produced by shrinkage during drying, when unconsolidated
mud is exposed to the air. They may be preserved when the mud is then covered by a new
depositional unit. They are a useful facing criterion.

UNCONFORMITIES
Unconformities are depositional surfaces representing a (time) gap in the depositional
history. Three types are commonly distinguished:
(a) Angular unconformity: layers below the unconformity are not parallel to the
unconformity. Since the layers above are bound to be parallel to the unconformity this results
in an angular discordance. Angular unconformities are easily recognized in undeformed
terrains, but may be obscured by deformation, which may tend to reduce the angular
discordance, and may also be confused with faults (see below).
(b) Disconformity (parallel unconformity): layers above and below are parallel.

EPSC 203 (2012) p. 26

(c) Non-conformity: the rocks beneath the unconformity are metamorphic or igneous (but the
contact is depositional).
Recognition of unconformities
A major problem is to distinguish unconformities from tectonic (faulted) contacts. The
following are some useful characteristics:
(i) Basal conglomerates (but beware of fault breccias). The beginning of a new phase of
deposition following prolonged periods of erosion is commonly marked by conglomerates coarse-grained sedimentary rocks consisting of generally rounded fragments. Commonly
these fragments are derived from the rocks that underlie the unconformity, although this may
not always be the case. It is rare for conglomerates to occur everywhere at the basal contact they fill depressions in the surface on which the new sequence is deposited.
(ii) Radiometric or paleontological dating. Under favourable circumstances this will
indicate clearly that the overlying rocks are substantially younger than those beneath.
(iii) Weathering of underlying material (and soil horizons). If there was prolonged erosion
before deposition of the younger succession, there may have been soils and subsoils
developed on the old surface. Soils, if preserved, are commonly very thin due to compaction,
but a deep weathering profile may be preserved with close to its original thickness. Care
must be exercised in the use of this criterion, because it is possible that juxtaposition of two
chemically different sequences by faulting will give rise to alteration at the contact during
metamorphism.
(iv) Contrasts in structural and/or metamorphic history. This may be expected at an
unconformity, with the younger rocks having a generally simpler history. This criterion,
however, also applies to many faults.
SOFT-SEDIMENT FOLDS
The term "soft-sediment folds" refers to folds produced in sedimentary rocks before
they were consolidated, i.e. before they were lithified by the cementation processes occurring
during diagenesis. The term "non-tectonic" folds is sometimes used, but is not favoured,
because soft-sediment folds may be developed as a result of tectonic events (such as the
riding of a thrust sheet over and within unconsolidated sediments).
(a) Compaction folds

EPSC 203 (2012) p. 27

Compaction folds are folds produced
by differential compaction of sediment during
lithification (Fig. 28), and are commonly
accentuated by initial draping effects, e.g.,
Figure 28
sediment-water interfaces dip away from
structures such as reefs. The recognition of
compaction folds is generally not a serious problem, because of their clear relationship to the
structures responsible for their existence, and because of their irregular development (spacing
& wavelength).
(b) Convolute folds
Convolute folds are folds in which
amplitude increases upwards from a flat
basal layer. There is little doubt that these
types of fold are produced during the Figure 29 Convolute folds developed in a
dewatering
process
associated
with lower sand layer in a graded bed.
lithification. They mark the pathways by
which water moved upwards out of the rock body. This type of origin is supported by the
existence of breached convolute folds in some instances. It is probable that the sites at which
convolute folds develop are determined by initial irregularities such as ripples (Fig. 29).
The recognition of convolute folds is not a problem since they have a very
characteristic style, and they are confined to certain horizons in the sedimentary succession.
(c) Slump folds
Slump folds are produced by the slumping of sediments that are not consolidated, but
are sufficiently compacted to maintain their cohesion during the slumping process. Such folds
occur quite commonly on submarine slopes on which the rate of sedimentation is high.
Slumping may be initiated by such things as earthquakes, regional uplifts, or simply by the
building up of the sedimentary succession into an unstable configuration. They need not,
then, necessarily be "non-tectonic" folds. Some slump folds are believed to have been
generated in unconsolidated sediments which were themselves moving as large masses
down submarine slopes.
The recognition of slump folds, and their distinction from "hard-rock" folds may be
extremely difficult. The most useful criteria for their recognition are:
(i) The presence of primary features exhibiting little evidence of the degree of
deformation evidenced by the folds; e.g., such things as undeformed animal burrows
or undeformed fossils. Where such features exist it is commonly possible to argue
that hard-rock deformation of an intensity sufficient to produce the observed folds
would not have allowed preservation of the features in such an undeformed state.

EPSC 203 (2012) p. 28

(ii) The presence of folds with very different attitudes. This criterion has to be used
with great care since a multiply deformed terrain will also yield folds with different
attitudes but of the same age. This is not, however, a common feature in the first set
of folds, if these folds are of hard-rock origin. There is, however, no necessity that
slump folds have very variable attitudes; many examples exist in which their attitudes
are consistent over large regions, reflecting a regional slope attitude.
(iii) Evidence of brittle failure, leading ultimately to the detachment of the limbs of folds.
Where present, this feature is one of the best indicators of soft-rock deformation. This
is because under hard-rock conditions it is unlikely that the mechanical contrasts
between two lithological types could be sufficiently large that one type (the limb) could
be undergoing brittle failure while the other (represented by the material around the
limb) was sufficiently weak to be able to flow in to fill the space created. Local
detachment of limbs does occur during hard-rock deformation, but in some slumped
terrains limbs have moved a long way since detachment, or are completely broken up.

DIAPIRISM
Diapirism is the gravity-induced piercing of an overlying rock unit by a less dense one
that underlies it. It may occur purely as a result of the history of sedimentation. The
phenomenon is particularly commonly observed with salt layers, because salt has a very low
porosity when formed. As a result it compacts little during lithification and becomes less
dense than the overlying sediments (siltstones, sandstones, etc.). If the density contrasts
become sufficient to overcome the strength of the overlying rocks, the salt begins to flow up
through the overlying rock, deforming it in the process. The site at which flow begins is
probably determined by irregularities on the salt surface.

The resulting intrusives of salt may

be circular or oval in plan, and may in some
cases break the surface. The structures they
produce are extremely important as oil traps
in many petroleum-producing regions (Fig.
30).
Figure 30
SEDIMENTARY FABRIC
Fabric = the general configuration of geometric elements in a rock. The term is used
to describe geometric elements on the field, hand-specimen or microscopic scale. In this
discussion we are concerned only with the microscopic fabric. For example, the fabric could
consist of randomly oriented grain boundaries or boundaries with a general (statistical)
preferred orientation. The "grain boundary" is one example of a "fabric element".

EPSC 203 (2012) p. 29

Sediments almost always have some preferred orientation of their fabric elements.
Common examples are:
(i) platy minerals settle parallel to the paleohorizontal, and are rotated into the
horizontal during compaction.
(ii) elongate mineral grains may be oriented parallel to the current direction.
(iii) under some circumstances, platy minerals are stacked so that they dip up-current
= "current imbrication".

EPSC 203 (2012) p. 30

PRIMARY IGNEOUS FEATURES
(a) Lava flows
Lava flows are tabular to lensoid in form, depending on their viscosity, and on the
configuration of the substrate onto which they are extruded; i.e., they may fill valleys. They
may be up to 100 m thick and may cover areas of up to several hundred km2. Generally,
lavas poor in SiO2 (such as basalt) are less viscous than those rich in SiO2 (such as rhyolite)
so that they have more tendency to form tabular flows. However, even the flows with lowest
viscosity - those extruded at high temperature, with low SiO2 and low gas content - develop
chilled rinds that constrain their flow, so that the flows are markedly convex upwards, at least
near their extremities.
Small-scale features
(i) Ropy lava (pahehe) and blocky lava (a). These are two terms applied to the surface
features of subaerial lava flows. Ropy lavas have smooth, generally glassy surfaces which
are thrown into folds and wrinkles by the moving lava underneath them. Blocky lava flows
have a rubble-strewn surface, the rubble being pieces of the lava flow that have broken up
due to the flow. Ropy-lava flows can be observed to change into blocky flows with increasing
flow rate.
(ii) Breccia. A volcanic breccia is any rock that consists of broken pieces in a volcanic matrix.
Commonly the broken pieces are themselves volcanic, and they may originate from the same
magma as the matrix. Alternatively the blocks may be exotic, picked up by the lava flow on
the way up, on its way from the vent to where it is, or thrown into the flow by explosive
eruptions.
(iii) Vesicles & amygdules (var. amygdales).
Amygdules = crystal-filled gas bubbles.

Vesicles = unfilled gas bubbles in lavas.

(iv) Pillow lavas = lavas with the morphology of pillows. They are generally regarded as
forming only underwater, where the efficient chilling characteristics of water produce a strong
chilled margin to the still molten flow. Many so-called "pillow lavas" are actually flow tubes.
Pillows commonly exhibit radial cracks and pronounced differences between core and rims.
Some pillows are hollow and carry internally layered flows.
(v) Columnar jointing = regularly distributed prismatic jointing which allows the rock to be
broken out into columns. Such jointing is common in thicker flows. It is due to shrinkage
during cooling of the flow.
The attitudes of the columns may provide a direct measure of the attitudes of the poles
to the surfaces of lava flows. There is, however, some departure from perfect orthogonality in
places where fluids have gained access to the flows via fissures, so that some care must be
exercised in using this method.

EPSC 203 (2012) p. 31

(vi) Flow banding. Flow banding is compositional banding in a flow, due to the segregation of
particular mineral species or grain sizes by fluid-dynamic processes. Again it may provide a
measure of flow attitudes.
(vii) Cumulate layering. Cumulate layers are produced by the settling of early-formed mineral
grains to the base of the still-molten lava flow. Since they are produced by settling, they may
be treated like sedimentary strata, providing a direct indication of the paleo-horizontal.
Facing criteria:
The following are the most useful facing criteria in ancient lava-flows:

(i) Form of pillow lavas: convex upwards, tongues protruding from

bases (Fig. 31).

Figure 31

(ii) Distribution of vesicles/amygdules: concentrated towards

the top of the flow (Fig. 32).
Figure 32

(iii) Filling of fissures in the top of a flow by

overlying sediments or another flow (Fig. 33)

Figure 33 Flow overlain by sandstone.

(iv) Differential chilling of the top and bottom of the flow, and displacement of the
coarsest-grained part of the flow towards the base of the flow.

EPSC 203 (2012) p. 32

(v) Location of cumulate zones (Fig. 34).

Figure 34

PYROCLASTIC ROCKS
Pyroclastic rocks are sedimentary rocks produced by explosive volcanic eruptions. A
general term for the products of pyroclastic eruption is "tephra". The "clasts" produced by
volcanic eruptions are classified according to grain size into :
(a) bombs and blocks, with diameter > 32 mm. "Bombs" were molten when ejected
and consequently have twisted or aerodynamically compatible forms. "Blocks" are
angular fragments of rock, which were solid before the eruption.
(b) lapilli; d 4-32 mm. Lapilli may be fragments of rock or glass, or individual crystals.
(c) ash; d < 4 mm
Some terms used for pyroclastic rocks:
"tephra" = any ejecta of pyroclastic origin
"agglomerate" = a deposit of volcanic bombs, although it is also variously used to refer to
volcanic conglomerates and volcanic breccias, and should be defined in context.
"tuff" = lithified ash
"lapilli tuff" = lithified lapilli-bearing pyroclastic deposit
"hyaloclastite" = "aquagene tuff" = clastic rock produced by the fragmentation of (glassy)
volcanic rocks on contact with water.
"ignimbrite" = deposit from a pyroclastic flow
"welded tuff" = sub-aerial tuff in which the fragments were still sufficiently warm when they
settled to weld together and form a single cooling unit like that of a flow. They may have such
features as columnar jointing.

EPSC 203 (2012) p. 33

Facing criteria for pyroclastic rocks are like those for clastic sediments. It is worth
noting, however, that air-fall deposits, and even the deposits from some ignimbrites, may
exhibit bedding attitudes that are far from horizontal - reflecting the slope of the surface on
which they were deposited. In addition, crystal tuffs, because of the marked density contrasts
between the mineral species that comprise them, may exhibit far from normal grading
characteristics.

INTRUSIVE BODIES
(a) Tabular Bodies
"dykes" (dikes) = cross-cutting (discordant) tabular bodies
"sills" = concordant tabular bodies
Dykes may exhibit layering and compositional zoning parallel to their walls, usually
symmetrical about their centres, due to fluid-dynamic sorting during intrusion and/or multiple
phases of intrusion. Sills may exhibit internal layering for the same reasons, and commonly
have well-developed cumulate layering. Some of the larger sills and dykes also exhibit other
sedimentary features such as cross bedding.
Criteria to distinguish sills from flows:
(i) symmetrical character of chilling features in sills
(ii) "baking" of over- and underlying rocks
(iii) apophyses of intrusive material in the overlying rock
(iv) imperfect concordance

Special tabular forms:

"cone-sheets" = conical sheets, convergent downwards
"ring-dykes" = sheets with circular-cylindrical form, or conical with a mild downward
divergence.
"laccoliths" = sills with upper contacts that are convex upwards
"lopoliths" = sills with lower contacts that are convex downwards
(b) Domal bodies

EPSC 203 (2012) p. 34

Domal intrusive bodies are distinguished as "stocks" if their area in horizontal section
is less than about 100 km2, and "batholiths" if they are larger. To some extent this
classification is dependent on the depth of erosion, since many "stocks" are connected to
"batholiths" at depth.
If batholiths are traced to depth in the crust it becomes very difficult to distinguish them
from the "country rocks" into which they were intruded. This is because the depth, and the
igneous intrusives, raise the temperatures of the country rocks to such a degree that they
have an appearance very similar to that of the intrusive itself. Consequently the overall form of
most batholiths is unknown, and perhaps even undefinable. Some batholiths at higher levels,
however, are known to have planar, horizontal bases. They have spread out along horizons
in the crust. They are then in a sense huge laccoliths.
The tops of batholiths may display evidence of the mode of intrusion of the pluton.
This mode may be by "stoping" ( = the breaking off and assimilation of pieces of country rock),
forcible intrusion ( = the bending of country rock out of the way), or by gradual transformation
of the country rock into a rock that resembles the intrusive material. This latter mode is more
important at greater depths, where temperatures are high, but all three modes may occur
together in a given intrusion.

EPSC 203 (2012) p. 35

TECTONIC FEATURES
"folds" = undulations in previously plane surfaces
"joints" = discrete planar fractures along which there has been little or no relative movement
"faults" = planar fractures along which there has been appreciable movement
"foliation" =

(a) any planar feature

(b) any planar feature of tectonic origin
(c) compositional banding of tectonic origin

"lineation" = any linear feature

"cleavage" = tendency of a rock to break along preferred planes
"penetrative" = developed on a sufficiently fine scale that there are no domains visible in
which the feature is not present; whether a feature is penetrative is dependent on the scale of
the observation
"slaty cleavage" = penetrative (on hand-specimen scale) cleavage in a rock whose grain size
is too small for individual grains to be seen with the unaided eye
"schistosity" = penetrative cleavage in rock with coarser grain size
"fracture cleavage" or "spaced cleavage" = non-penetrative cleavage

EPSC 203 (2012) p. 36

FAULTS
Faults are discrete planar features along which there has been movement. The
movement results in offset of features that were previously continuous across the fault. The
amount of offset of a particular feature is referred to as the displacement of the fault. The
displacement is not a fundamental feature of the fault, since it is dependent on the attitude of
the feature displaced, relative to the attitude of the fault, and on the attitude of the section
(map or outcrop) in which the displacement is observed. It is, therefore, used in only a
colloquial sense.

Figure 35. Perspective view of fault cutting bedding plane. DS = dip separation, SS =
strike separation, HS = horizontal separation in dip-bearing direction.
A more precise way of specifying the effect of a fault on a previously continuous

EPSC 203 (2012) p. 37

feature is to describe the separation of the feature by the fault. The separation of a feature,
in a given direction, is the distance in that direction that separates the feature on one side of
the fault from its equivalent on the other side. One may therefore define several
displacements for a given feature. e.g., a horizontal north-south separation, a vertical
separation. The strike separation is the separation along the strike of the fault. Note that, if
the attitude of the feature being offset is known, only one separation is necessary fully to
specify the effect that the fault has on the feature (Fig. 35).
The separation, however, is also not a fundamental feature of the fault. It is dependent
on the relative attitudes of the fault and the feature offset.

Figure 36. Perspective view of a fault offsetting two planes. The strike separations (SS1, SS2
are shown for each plane. AB is the net slip on the fault plane.

The fundamental feature of the fault is its slip which is the actual motion on the fault
plane. It is a line, with an orientation and a magnitude, lying in the fault plane. The net slip
may be resolved into two components on the fault plane, such as the "strike slip" and the "dip
slip" (Fig. 36, 37).

EPSC 203 (2012) p. 38

Figure 37. Sketch in the fault-plane surface of Fig. 36. NS

is the net slip, DS is the dip slip, SS is the strike slip.
Faults are classified according to the
sense of the strike-slip and dip-slip
components.

Dip-slip: The fault block that lies above the

fault-plane is referred to as the "hanging
wall", and that beneath it is referred to as the
"footwall". A fault in which the hanging wall
has moved downwards relative to the Figure 38. Normal (a) and reverse (b) dip-slip
footwall is a normal fault. One in which the faults, viewed along their strike lines.
hanging wall has moved relatively upwards
is known as a reverse fault (Fig. 38).

EPSC 203 (2012) p. 39

Strike-slip: A fault in which the block on the
opposite side from the observer has moved
to the right is a dextral or right-lateral fault.
A fault in which the block on the opposite
side has moved to the left is a sinistral or
left-lateral fault (Fig. 39).
Many faults have either an overwhelming strike-slip or an overwhelming dipslip component. The former type is known Figure 39. Dextral (a) and sinistral (b) strikeslip faults, in plan view.
as a wrench fault.
SOLUTION OF FAULT PROBLEMS
A recommended method for the solution of fault problems is as follows:
(a) Project all structures displaced by the fault into the plane of the fault; i.e., determine the
attitudes of their intersections with the fault plane and their relative positions in the fault plane.

(b) Draw a plan of the fault plane, i.e., a section in the fault plane. Note that the information
you must use to do this is the pitches of the various structural elements in the fault plane.
(c) If only one structure was displaced it is immediately clear that there is no unique solution
for the net slip. If more than one structure was displaced, determine the piercing point for the
line of intersection (i.e., the point at which this line of intersection pierces the fault plane) of
two structures, for the block on each side of the fault plane. The line joining these two
piercing points is the net slip.

EPSC 203 (2012) p. 40

As an example, consider
the fault whose surface relationships are shown in Fig. 40.
The plan diagram of Fig. 40
may be used to prepare a
stereogram (Fig. 41). From this,
the pitch of the bed and the dyke
in the fault surface may be
determined.
Figure 40. Map of east-striking fault cutting a bedding
contact striking 060 and a dyke striking 140.

EPSC 203 (2012) p. 41

These pitch angles may be used
to construct a view in the fault
plane (Fig. 42). The pitch of PQ
may be converted into a trend
and plunge on a stereogram.
The magnitude of PQ may be
measured directly.

Figure 41. Stereogram of fault, bedding and dyke from

Fig. 40. The pitch of bedding in the fault is 1 (= 20oW).
The pitch of the dyke in the fault is 2 (= 102oW).

Figure 42. Sketch in fault plane of Fig. 40. PQ is the net slip.

EPSC 203 (2012) p. 42

Dip Slip
"Normal fault" = hanging-wall down.
"Reverse fault" = hanging-wall up
The dip component of slip on this fault (RQ) has moved the hanging wall up. This is
therefore a reverse fault.
Strike Slip
"Dextral fault" = block away from you has moved to the right.
"Sinistral fault" = block away from you has moved to the left.
We are viewing the fault plane from the south (hanging wall). The strike component of
slip (PR) displaces the footwall to the left. This is therefore a sinistral fault.

FAULT SOLUTION ON MAPS

When working with representations of faults on maps, it is desirable to solve the
problems directly on the maps. This provides a much clearer visualization of the problems
than can be derived from extraction of the data onto stereograms. The simplest way to treat
the data directly on the map is through the preparation of structure-contour maps of each of
the planar features represented on the map. Structure contours, as the name implies, are
contour lines that depict the structure of a surface, in the same way that topographic contours
depict the relief of Earth's surface. For planar, undistorted surfaces, such as those we have
discussed so far, structure-contour maps consist of equally-spaced lines parallel to the strike
of the surface ('strike lines'). As we have seen, steeply-dipping surfaces have closely-spaced
strike lines, shallowly-dipping surfaces have more broadly-spaced strike lines. These strike
lines may be used to construct complete subsurface (and supersurface) maps of regions of
great complexity.
In Figure 43, strike lines have been used to construct the horizontal projections of lines
of intersection of the sedimentary contact with the fault (S-F), the dyke with the fault (D-F) and
the sedimentary contact with the dyke (S-D), for each fault block. S-D, for example, must
pass through the intersections of strike lines for the sedimentary contact and the dyke that are
at the same elevation. For each fault block, the intersection of these three lines is the point at
which the line of intersection between the sedimentary contact and the dyke meets the fault.
These points are shown as A and B on the figure, and may be used to determine the net slip
on the fault. Note that in fact only two of these lines of intersection were needed to locate A
and B; the three were constructed to provide a check.

EPSC 203 (2012) p. 43

Figure 43. Map of a fault offsetting a sandstone/shale contact and a dyke. Strike lines are
shown for all planar features (solid lines), as well as the horizontal projections of the lines of
intersection between features (dashed lines).

EPSC 203 (2012) p. 44

STRESS

Figure 44
A body in equilibrium under a system of forces is said to be in a "state of stress". The
system of forces is the "stress system". "Strain" is the response of a body to a system of
stress. The system of forces acting on a minute cube in a body may be depicted as shown. It
consists of "normal" forces (fii), perpendicular to the surface on which they act, and
"tangential" or "shear" forces, parallel to the surface on which they act (Fig. 44).
One may consider the limiting case, in which the minute cube becomes infinitesimally
small. In this case the forces too become vanishingly small. However, the ratio of the force
to the area on which it is acting remains a finite quantity. It is the ratios that are referred to as
the stress components. They are symbolised by ij rather than fij, and they are defined at a
point as the stress field at that point. The stress field therefore has 9 components at any
given point, and the magnitudes of the components are determined by the orientation of the
Cartesian coordinate system in which the stress field is being described. Stress is an
example of a "second-rank tensor" or, more commonly, a "tensor". That is, it is a variable
array with two subscripts and nine components, the values of which change according to a
specific set of rules if the coordinate system in which they are described is changed. A
"vector" is a "first-rank tensor".
For any tensor, and more specifically for the stress tensor, it is always possible to
choose a Cartesian coordinate system in which to describe the tensor, in which all but the

EPSC 203 (2012) p. 45

normal components of stress disappear (compare with vectors, for which one may always find
a coordinate system in which all but one of the components disappears). The axes of this
coordinate system are referred to as the "principal axes" for the stress. In this coordinate
system the stress tensor may be given as

1 0 0

0 2 0

0 0
3

rather than

11 12 13

21 22 23

31 32 33

i are the "principal stresses". By convention they are generally chosen so that 1 >
2 > 3, where a compressive stress is positive and a tensional stress is negative.
Note that, because of our choice of coordinate system, there are no shearing stresses on
planes that are parallel to the principal stress axes. This does not mean that there are no
shearing stresses in the system; any plane at an angle to the principal stress axes will
experience shearing stress.
The "mean" stress is given by:

+ +
= 1 2 3
3

Using this mean stress we may define an "isotropic" stress tensor, which is a stress tensor in
which the magnitudes of all principal stresses are the same:

0 0

0
0

0 0
In addition, we may subtract this isotropic tensor from our original tensor, to give the
"deviatoric stress tensor":

EPSC 203 (2012) p. 46

( 1 - )
0
0

0 ( 2 - )
0

0
0 ( 3 - )

The isotropic stress tensor is similar in character to a hydrostatic pressure ; stresses

are the same in all directions and there is no shearing stress in any direction. It is therefore
responsible only for changes in the volume of rocks, not for changes in their shape. This
mean stress is a result both of the burden of rocks overlying those under consideration, which
gives rise to a "lithostatic pressure", and to externally applied or internally generated stresses.
It may therefore differ in magnitude from the lithostatic pressure, although probably not a
great deal.
The deviatoric stress is responsible for the change in shape of rocks in a stress field,
and leaves the clearest evidence of its existence in the form of structures.

MOHR STRESS CIRCLE

As mentioned above, a plane oriented at any angle to the principal stress axes will
experience shearing stress. The Mohr stress circle is a mathematical gimmick by which the
magnitudes of these stresses may be evaluated if the principal stresses are known.
It is derived from expressions for the normal and tangential stress on any plane at an
angle to 1 (Fig. 45).
normal stress: = 1 sin sin
where 1 sin is the component of 1 normal to the
plane of interest, and sin is an adjustment for the
increase in area over which the force is acting (i.e.
we are interested in force/unit area, not force itself).
tangential stress: = 1 cos sin
Figure 45. Resolution of principal
If we then add another principal stress, 2, stresses into normal and shear
normal to 1, and treat it in the same way, we arrive stresses on an arbitrarily oriented
plane.
at the expressions:
= 1 sin2 + 2 cos2

EPSC 203 (2012) p. 47

= (1-2) sin cos
substituting sin 2 = 2 sin
cos ;
cos 2 = cos2 - sin2 ;
sin2 + cos2 = 1, we get:
=(1+2)-(1-2)cos 2
=(1-2) sin 2
These are the equations of
the Mohr circle (Fig. 46).
Note that the angles
are measured from 2 on
the circle, whereas in the
rock they are measured
from 1. This is because
Figure 46. Mohr stress circle, showing relationship compressive stress was
between normal stress and shear stress on planes oriented defined as positive.
at angles to 1.
NAVIER-COULOMB FAILURE CRITERION

The
NavierCoulomb
failure
criterion is a criterion
for failure of a material
on a given plane, in
terms of the normal
and shearing stresses
on the plane, the shear
strength
of
the
material, and the angle
of internal friction.

It may be written:
= S + tan
where is the angle of
internal friction, and S Figure 47. Navier-Coulomb failure criterion, and Mohr circle.
is the shear strength.
The criterion may be stated: "When ( - tan ) exceeds S on any plane, the material will fail

EPSC 203 (2012) p. 48

on that plane".
On a Mohr circle, the failure criterion appears as a line (Fig. 47). For the circle shown,
representing a particular homogeneous stress field, failure will occur on a plane oriented at
to 1, since on that plane:
- tan = S
From the geometry of the diagram it is clear that
2 = 90o - , i.e. = 45o-/2,
where is the angle at which failure occurs.
Note that on Figure 47 the Mohr circle depicted has intercepts 1 and 3, and therefore
represents planes that contain the axis 2. This is because planes containing axis 2 always
have the largest shear stresses for a given value of the normal stress. The plane of failure is
therefore generally likely to contain the axis 2 if the rock is isotropic.

A Plane of Weakness

A
plane
of
weakness in the rock
will be represented by a
failure criterion with
different parameters:
lower in most cases.
On the Mohr circle the
new failure criterion will
appear as on Fig. 48.

Figure 48. Orientations of planes for which failure will occur at

deviatoric stresses lower than those for the failure criterion of the
rock, if the planes have lower internal friction.
If the plane of weakness is oriented at any angle within the shaded area, failure will
occur on that plane rather than at the theoretically predicted angle on some new plane cutting
through the rock.

EPSC 203 (2012) p. 49

STRAIN
Strain is the response of a body to a stress. This response consists of a volume
change and/or a change in shape. Bulk translations are usually excluded from the definition.
Strain at a point in a body may be described in terms of nine components, just like the
stress. Consider a point (x1, x2, x3). After strain this same point has new co-ordinates x1', x2',
x3'. Each xi' is a function of all xi.
e.g., x1' = x1 + 11x1 + 12x2 + 13x3 ,

where ij are the components of strain.

Thus we have the strain tensor:

11 12 13

21 22 23

31 32 33
The condition for irrotational strain is that ij = ji when i j.
Thus, we may break the strain tensor into a rotational and an irrotational part:
strain

rotational

irrotational

0 12 - 21 13 - 31 2 11 12 + 21 13 + 31
11 12 13

2 21 22 23 = 21 - 12
0 23 - 32 + 21 + 12
2 22 23 + 32

0
+
+
2
33
31 32 33 31 13 32 23
31 13 32 23
When treating strain it is generally necessary to consider only the irrotational part. The
rotation is removed, since it involves an assumption of an (arbitrary) reference frame. We
may, as for stress, choose a set of Cartesian coordinates so that all off-diagonal components
disappear, and the irrotational strain tensor becomes:

0
0
11

0 22
0

0
0 33

which is normally simplified to:

EPSC 203 (2012) p. 50

1 0 0

0 2 0

0 0
3

The axes so chosen are the "principal axes of strain".

Conventionally, 1 > 2 > 3.
Homogeneous strain = strain for which i are constant throughout the body. Any strain may
be considered homogeneous if a sufficiently small element of the body is considered. Thus,
we normally treat only homogeneous strain mathematically.
Isotropic strain = strain for which all i are equal.
Properties of homogeneous strain
The following properties of homogeneous strain may all be proved quite simply:
(a) Straight lines are converted to straight lines. Thus planes remain planar.
(b) Circles are transformed to ellipses. Spheres are transformed to ellipsoids. This is a very
important property.
i.e., consider the sphere x12 + x22 + x32 = 1 (sphere of unit radius)
During strain xi xi' where xi' = xi + ixi = (1+i) xi
!

i.e. xi =

xi
1+ i

Thus, the equation for the body the sphere has become is:
!2

!2

!2

x1
+ x 2 2 + x3 2 = 1
2
(1+ 1 ) (1+ 2 ) (1+ 3 )
i.e., an ellipsoid with semi-axes (1+1), (1+2), (1+3).
This ellipsoid is the strain ellipsoid, and it provides the simplest representation of the
nature of strain.

EPSC 203 (2012) p. 51

(c) There are always three orthogonal lines that remain orthogonal after the strain. They may
rotate during the deformation, while remaining orthogonal, in which case the strain is
"rotational", or they may remain in fixed orientations, in which case the strain is "irrotational".
Strain may always be treated as irrotational, simply by regarding the strain of a body from a
rotating reference frame.
Longitudinal and Tangential (Shear) Strain
Both types of strain may be defined for a given direction in a material, at a given point.
Longitudinal strain = the change in length of a unit length in that direction.
Shear strain = the change in a right-angle, one of whose sides is the direction in which the
strain is being measured.

For a given direction in a body there are

two tangential strains, measured perpendicular to
each other (Fig. 49). In Figure 49, B has been kept
fixed, so that all changes are measured relative to
it, and the shear strains are depicted for that
direction.
OA, OB and OC were orthogonal. Angle BOA is
now < 90o. (90-BOA') is the "angular shear strain"
in plane BOA.

Figure 49

Two-Dimensional Strain
In two dimensions a circle is transformed into an ellipse during a homogeneous strain
(Fig. 50). The radius of the ellipse, in any direction, has length (1+) where is the
longitudinal strain or "stretch" in that direction.
The shear strain is given by the angle = change in a right-angle (OA and PQ were
perpendicular before deformation). Shear strain is more commonly quoted as = tan .

EPSC 203 (2012) p. 52

Figure 50
Note that is zero parallel to either of the principal axes. Thus there is no shear strain
in the principal planes of strain (cf. no tangential stress in principal planes of stress).

Reciprocal Quadratic Strain

If we establish the following definitions:
= (1+)2; ' = 1/; ' = /
it may be shown (using straightforward trigonometry) that ' and ', which are the "reciprocal
quadratic strains" (longitudinal and shear respectively), obey the following relationships:
' = 1' cos2 + 2' sin2;
' = (2' - 1') sin cos
These may be used to set up a Mohr strain circle, used in the same way as the stress
circle (Fig. 51).

EPSC 203 (2012) p. 53

Figure 51

Strain & Volume Change

The volume of an ellipsoid is given by 4abc/3; that of a sphere is 4/3. A unit sphere, with
volume 4/3 therefore becomes an ellipsoid with volume 4(1+1)(1+2)(1+3)/3. The change
in volume of a unit volume is therefore given by (1+1)(1+2)(1+3) - 1. This is known as the
dilatation, and is approximately 1 + 2 + 3.

DETERMINATION OF STRAIN
Although strain is a three-dimensional feature, the strain in a given body may be
exactly determined by measuring the two-dimensional strain in three mutually orthogonal
sections. The determination of strain is therefore a two-dimensional problem. There are
many different methods available:

(a) Ramsay and Dunnet's method for deformed pebbles.

EPSC 203 (2012) p. 54

This method is based on the assumption that the pebbles were randomly oriented
before the strain. Although this is clearly a simplification, the assumption can be checked
(see below under Robin's method).
If all the pebbles were spherical before strain, then the strain will convert all the
pebbles into ellipsoids, with the same eccentricities, and all with their long axes parallel
(assuming all pebbles have the same mechanical properties). The orientation and ratios of
the strain ellipse may be read off directly.
If all the pebbles were
ellipsoidal to begin with, with the
same initial axial ratio, but were
randomly oriented, then the
pebbles whose long axes
coincide with the long axis of
strain will develop extreme
Figure 52
eccentricities.
The pebbles
oriented with their long axes
perpendicular to that of the strain will develop much smaller eccentricities, but with the same
axes (Fig. 52).
Pebbles oriented at some
angle to the ellipse of strain will
develop
intermediate
eccentricities, with long axes
oriented at some angle to those of
the strain. It is possible to show
that the relationship between the
orientation of the long axis of the
final ellipse and its eccentricity
(axial ratio) has the form shown on
Fig. 53.
On this figure:
Ri = initial ratio of pebbles
RS = ratio for the strain
= angle between long axis of
deformed pebble and an arbitrary
reference direction.
Figure 53
A diagram such as this
may be prepared for any set of
randomly oriented pebbles. From it, the direction and ratios of the strain ellipse may be
determined, by measuring the angle at which the eccentricities are greatest, and measuring

EPSC 203 (2012) p. 55

the maximum and minimum eccentricities.
It is unlikely that in practice all pebbles would have had the same initial ratios,
especially if the conglomerate in which they occur is polymictic. If there is a range of initial
ratios, pebbles with smaller initial ratios will plot within the field outlined by those with the
largest initial ratios on the graph.
The principal problems with the method are the assumption of random initial
orientation (see below) and the likelihood that the pebbles will have experienced less strain
than the (weaker) matrix in which they lie. The method can therefore yield only a minimum
estimate of the strain. It is also of course of relatively limited utility, since it requires the
presence of initially ellipsoidal markers. It has however been used successfully in many
conglomerates and oolitic limestones.
(b) Wellman's method.
Wellman's method relies on the
recognition of original (before strain)
right-angles in the deformed rock. It uses
the geometric property that the angle
subtended by the diameter of a circle on its
circumference is a right-angle.
If we
consider a set of such subtended
right-angles, subjected to strain, they are
converted into angles other than 90o, all of
which are subtended on an ellipse, which
has the same attitudes and ratios as the strain ellipse (Fig. 54).

Figure 54

The problem therefore reduces to the

recognition of the previous right-angles and
constructing the ellipse on which the apex of
each of these angles lies when subtended
from a common line. The choice of the line
is somewhat arbitrary, since it determines
only the scale and position of the ellipse
(Fig. 55).
Figure 55
(c) Mohr-circle method.
The Mohr circle method is also based on the recognition of original right-angles in the
deformed rock. Unlike Wellman's method, however, only two such (former) right-angles need
be known, provided they are oriented differently.

EPSC 203 (2012) p. 56

The problem is approached
by constructing the Mohr circle for
the strain, by using the property of
the Mohr strain circle that the
angle between the ' axis and the
line joining the origin to a point
with given values for ' and '
(which represents a line in the
strained rock with a particular
direction) is equal to the angular
shear strain in that direction (Fig.
56).
If the angular shear in two
different
directions
can
be
Figure 56
determined, the orientation and
ratios of the Mohr circle may be determined uniquely. The angular shear may be identified as
the change in angle of a former right-angle; hence the importance of the recognition of former
right-angles.
Procedure :

Figure 57

Identify two former right-angles,

determine the values of their
angular shear, and the sense of
angular shear.
Note that the
sense of angular shear is
dependent on which arm of the
former right angle is chosen as the
reference direction (Fig. 57). In
Fig. 57, the heavy lines are those
used as references, and the sense
of shear is considered positive for
a clockwise rotation of the other
arm of the former right angle, with

respect to the reference arm.

These two sets of reference-direction + angular shear data may now be used to
construct two lines emanating from the origin of a ' - ' graph, offset by the appropriate
angular shear angles from the ' axis (Fig. 58), and a Mohr circle (so far unattached to the
axes - on a separate sheet of paper, Fig. 59) on which are superimposed two radii, separated
by twice the angle of separation of the reference directions (since directions are doubled on
the Mohr circle).

EPSC 203 (2012) p. 57

Figure 58

A reference direction related to some external frame

(e.g., N) should also be included. The two diagrams may
now be superimposed to provide a unique fit compatible
with the data contained on each (Fig. 60). From this
superimposed diagram both the orientation and ratios of
the strain ellipse may be determined.

Figure 59

Figure 60
(d) Robin's method
This last method has the advantage of being useful for strain markers of any shape,
whether or not they have a preferred orientation or non-equant shapes to begin with.
Assuming a random orientation, perpendicular "diameters" of a large number of
markers will sum geometrically to 1, before deformation.

EPSC 203 (2012) p. 58

i.e.,

a1 a 2
a
x x ... x n = 1
b1 b2
bn

where n is large.
After strain,
n
a1 a 2
a (1 + 1 )
x x ... x n =
n
b1 b2
bn (1 + 2 )

provided the "diameters" were measured in the principal strain directions. These directions
are not known initially, but may be determined by an iterative process, in which the directions
for which the geometric sum is maximized are determined. The geometric sum then gives the
value of

(1 + 1 )n
(1 + 2 )n
To test for randomness of initial orientation, an Rf - (see Ramsay and Dunnet's
method) plot may be generated for the principal strains calculated. If the "tear-drop" field is
fully occupied, there was a random orientation of markers with different initial ratios. If only
segments of the field are occupied, there was a preferred orientation of the markers.
PROGRESSIVE STRAIN
The strain we observe in a rock is the total strain that has been experienced by the
rock during the deformation history. It may be considered to result from the superposition of
an infinite number of stages of infinitesimal strain, each represented by an infinitesimally
eccentric strain ellipsoid. The strain path, which is the sequence of infinitesimal strain
ellipsoids experienced by the rock, is commonly indeterminate. A major distinction may be
drawn between progressive strain paths for which the principal axes of infinitesimal strain
remain parallel to the same material lines within the body throughout the deformation, for
which the strain is referred to as coaxial, and progressive strain paths in which the material
lines that are parallel to the infinitesimal principal strain axes change as the deformation
progresses, for which the strain is referred to as non-coaxial. For strain in two dimensions
only (i.e., strain for which there is no stretching or shortening of lines in one direction in space,
so that the strain may be fully described by viewing it just in the plane containing the other two
orthogonal directions in space), which is known as plane strain, coaxial strain is referred to
as pure shear.

EPSC 203 (2012) p. 59

Pure shear results from the progressive superposition of infinitesimal strain ellipses all
of which are oriented in the same way with respect to the body being deformed. Each
progressive strain increment produces extension parallel to the long axis of the strain ellipse,
contraction parallel to the short axis, no extension at 45 to the principal axes, and smooth
variations of the longitudinal strain in directions intermediate between these directions (Fig.
61). The total strain is the result of many successive applications of infinitesimal strains like
that of Fig. 61. Because the principal axes remain parallel to specific material lines within the
body, these lines experience a simple strain history, either progressive elongation, in the case
of lines parallel to the long axis of the ellipse, or progressive contraction, in the case of lines
parallel to the short axis. Pure shear is the kind of progressive strain that would be
experienced in a homogeneous body subjected to an irrotational stress field with 2=0. Such
a stress field is not common in nature, so pure shear is also not common in nature. A
reasonably common irrotational stress field is, however, one in which 2=3. For a
homogeneous rock, such a stress field would also produce a coaxial strain, transforming a
sphere into a hamburger- or pancake-shaped body (a 'flattening' strain ellipsoid), depending
on the magnitude of 3.
Consider a pure shear. It is easy to show that material lines in a body subjected to a
progressive strain of this type may undergo an initial shortening, followed by lengthening,
during the deformation:

Figure 62

Figure 61. The sphere and the infinitesimal

strain ellipse (exaggerated eccentricity)
derived from it.

At any given time during the deformation, lines within 45o of (1+1) are lengthening,
and lines within 45o of (1+2) are shortening (Fig. 61). The net effect of the progressive strain

EPSC 203 (2012) p. 60

is, however, to rotate all material lines towards (1+1). This can be shown as follows:
Consider a line at an angle to (1+2) before strain.
tan = a/b
During strain: a (1+1)a, b (1+2)b
The situation before strain is as in Fig. 62.
After the strain (Fig. 63) has become ', where:

tan =

(1+ 1 ) a (1+ 1 )
. =
. tan
(1+ 2 ) b (1+ 2 )

Since

(1+ 1 )
>1
(1+ 2 )
(by definition), it follows that '>.
Thus, a material line that
began life by contracting (within
45o of 1+2) will rotate into an
extensional field.

Figure 63
We may divide the ellipse into fields in which lines have had different histories (Fig.
64). Similar divisions of material space would apply to sections through a flattening ellipsoid,
provided the strain was coaxial. The only difference would be that the angle 45 would be
replaced in the figure by the smaller angle 26.6 (which is the angle between the short axis of
the strain ellipsoid and the lines of no elongation for an infinitesimal perfect flattening ellipse;
1+2 = 1+3) and the boundary between fields B and C would move correspondingly. The
progressive strain history could be much more complex for a non-coaxial flattening strain.

EPSC 203 (2012) p. 61

Figure 64. In Field A, lines have contracted throughout their history.

In Field B they first contracted, then extended. In Field C they have
always extended.

The simplest type of non-coaxial strain is known as simple shear (Fig. 65). Simple
shear is also a plane strain (no strain in the third dimension; strain may be fully represented
by a strain ellipse in two dimensions). Strain is in response to a shear couple. There is no
deformation in the plane parallel to the shear couple. The strain pattern for simple shear may
be illustrated with the strain experienced by a deck of cards (Fig. 65c) . During the
deformation, the distance from edge to edge of the card deck remains the same (Fig. 65b);
the same is true of simple shear. The infinitesimal strain ellipse for simple shear always has
a long axis at 45 to the plane of the shear couple (Fig. 65a). In Figure 65a, a circle and the
strain ellipse after finite simple shear are shown. Note that the material line that was parallel
to the direction of infinitesimal maximum elongation at the start of the deformation (AA') has
now rotated towards the shear plane so that it is now no longer parallel to the direction of
maximum infinitesimal elongation. This is the defining characteristic of a non-coaxial strain.
The only material line that does not rotate during simple shear is the line DD', parallel to the
plane of the shear couple. This line also does not experience any elongation. The plane of
the shear couple, commonly called the flow plane, does not experience any deformation,
because there is no strain in the third dimension. With progressive strain, the total strain
ellipse becomes more eccentric, and its long axis rotates towards the flow plane. The amount
of strain in simple shear is commonly monitored through the angle (Fig. 65b).

EPSC 203 (2012) p. 62

Simple shear provides a reasonable first approximation to the strain history in many ductile
shear zones.

lo
e s ng
im a x
al is
s tr - i
ain nfi
e l ni t lip
se

C'
B'

lo
e s ng
im a x
a l is
str - i
a i n nf i
e l n itlip
se

EPSC 203 (2012) p. 63

C'

45
D

D'

D'
strain
ellipse

B
C

c
card-deck analogue
Figure 65. Simple Shear

B'

A'

EPSC 203 (2012) p. 64

BEHAVIOUR OF MATERIALS
Instantaneous Deformation
Solids, deformed by a deviatoric
stress, exhibit a stress - instantaneous strain
curve as in Fig. 66.
It may be divided into an elastic
segment, and a plastic segment.
ELASTIC:

(i) (linear relationship)

(ii) strain is recoverable

PLASTIC:

(i) strain is not recoverable

(ii) there is a larger strain for a
given stress increase than in
the elastic domain.

Figure 66

Y = YIELD POINT = the stress value

marking the boundary between elastic and plastic strain.
Note that the plastic strain curve still
has a slope > 0. Thus, in a stress history
producing the strain 1 (Fig. 67) if the stress
is released the elastic strain is lost and the
final, permanent strain is 2. If the material
were stressed again, however, it would
experience no plastic strain until the
threshold stress Y2 was exceeded, not Y1
as before.
Thus, the yield point has
increased.
This is known as "WORK
HARDENING".
"Ideal" plastic materials exhibit no
Figure 67
work hardening (slope = 0). Real materials
generally have a small positive slope, so that
the amount of strain that can be produced at
a given stress is finite. This can serve as a general (colloquial) definition of "plastic"
behaviour. In practice, however, the term "plastic" is used to refer to time-dependent strain
that occurs only above a yield point, as distinct from viscous strain (see below).
Elastic constants
The stress - strain dependence is generally written as = k, where k is a constant.

EPSC 203 (2012) p. 65

For normal stress and longitudinal strain: = E, where E is Young's modulus.
For shear stress and shear strain: = G, where G is the Shear or Rigidity modulus.
Generally, only the two moduli (or two equivalent moduli) are needed to provide a
complete description of the elastic response of a body.
Failure
Loss of cohesion (breakage) of the material could occur before the yield point was
exceeded (i.e., at lower stress). In this case the material is "BRITTLE", and the failure it
experiences is "BRITTLE FAILURE". If loss of cohesion occurs in the plastic domain there is
some permanent strain before failure, and the failure is "DUCTILE".
"DUCTILITY" is simply the tendency of a material to exhibit permanent strain without
loss of cohesion. It is sometimes defined as the percentage of strain before rupture, but is
more commonly used in a colloquial sense.

EFFECT OF PRESSURE AND TEMPERATURE

(a) Pressure
Increasing P can produce a minor increase in the yield point. It has a major effect on
the ULTIMATE STRENGTH (= maximum deviatoric stress a material can sustain without
failure). Thus, a brittle rock could become ductile with increase in confining pressure.
Typical curves for the effect of
increase in pressure are like those of Fig.
68.
(b) Temperature
Increasing T lowers the yield point
very effectively. It also raises the ultimate
strength.
Typical geotherms exhibit both
increasing P and T. All materials therefore
become more ductile with increasing depth
in Earth.
TIME-DEPENDENT STRAIN

Figure 68

EPSC 203 (2012) p. 66

If stress operates for a long period, the material will develop more permanent strain
than its instantaneous plastic strain. Even materials that are elastic under the given
conditions ( < Y) will exhibit permanent, time-dependent strain. Time-dependent strain is
known as creep.
The typical creep curve for a material deformed at a given deviatoric stress,
temperature and pressure is shown on Fig. 69.
On this figure, i = instantaneous strain, which may be elastic or elastic + plastic.

Primary
creep
is
generally
decelerating; i.e., the strain rate is
decreasing with time. In rare cases it is
accelerating. Its defining characteristic is
that it is recoverable, although the return to
unstrained state takes time.
Pseudo-viscous or secondary creep
is geologically the most significant, because
a large degree of strain can be achieved by
it if the stresses are operating for long
periods of time. Its defining characteristics Figure 69. I is the primary creep, II is the
are that it is linear, and is not recoverable.
secondary or "pseudo-viscous" creep and III is
the tertiary creep.
Tertiary or accelerating creep is
simply the precursor to rupture, and will not occur in materials that are being deformed at
stresses significantly less than their ultimate strength. Note, however, that it is not an
increase in stress that produces the rupture here. Rather it is a result of the material's having
experienced too much strain. "Ultimate strength" cannot therefore be rigorously defined
when creep is considered. The geological expression of failure on the creep curve is a ductile
fault, or a fault developed in a fold during the folding.
Equivalent Viscosity
For the pseudo-viscous creep we may write the expression

EPSC 203 (2012) p. 67

since

is constant for a given .

t

This is the flow law for a Newtonian viscous fluid, where is the VISCOSITY, and for a
fluid that is strictly "Newtonian" is independent of (but not of T). For rocks this is generally
not the case; i.e., is constant for a given , but the viscosity is a function of the stress
applied. The viscosity is also strongly temperature dependent.
The actual flow law obeyed by rocks varies with the microscopic strain mechanism,
which is dependent mainly on temperature and grain size. At low temperatures the viscosity
may be essentially independent of stress, and therefore approximately Newtonian, but at
higher temperatures ( > 400oC) the flow law for materials is generally of the form:

k
n t

where n is between 2 and 4, and k is a constant.

This equation may be rewritten:

1 n+1
=
t k
Creep obeying this relationship is known as "Power-law" or "Weertman" creep. For this creep
the equivalent viscosity is k-n and is strongly dependent on the stress.
The temperature dependence of viscosity, for both power-law and quasi-Newtonian
creep is of the form

k = k 0 exp

CT m
T

where C, and k0 are constant for the material and Tm is its melting temperature in Kelvin
degrees. This gives an exponentially decreasing viscosity with increasing temperature.
Viscosity may decrease by 1 or 2 orders of magnitude for a 100o temperature increase.

EPSC 203 (2012) p. 68

FOLDS AND FOLDING
(a) Folding of a Single Surface
The simplest type of folding of a surface is cylindrical folding, in which the surface
may be generated by moving a line parallel to itself through space. An alternative way in
which to define a cylindrical fold would be to say that it is a fold for which all planes tangent to
the fold surface contain a common linear (which has the same attitude as the line used to
generate the fold in the first definition).
Although many naturally formed folds are not cylindrical, it is conventional to treat them
as if they are. Many folds are approximately cylindrical, and if they are not they can be
divided into domains that are approximately so.

Figure 70

The attitude of the line common to all planes on the fold is referred to as the attitude of
the FOLD AXIS. Strictly speaking any line with that attitude is a fold axis, although the term is
generally used for a specific line (the hinge line - see below). The geometry of the fold is then
most simply represented in the plane perpendicular to the fold axis, which is referred to as the
PROFILE PLANE.
Within the fold profile one may define a CREST and a TROUGH for a fold. These are
the lowest and highest points on the fold. One may also define a crest line and a trough line
parallel to the axis. These features are, however, dependent on the external reference frame;
they are not intrinsic features of the fold (Fig. 70a).
The intrinsic features of the fold are based on its curvature. Thus, one may define
INFLECTION POINTS on the profile, at which the curvature is zero and the sense of
curvature is changing (e.g., from left-curving to right-curving) and HINGES, at which the
curvature is a maximum (Fig. 70b). Conventionally, a "fold" is the region between two

EPSC 203 (2012) p. 69

adjacent inflection points on the surface. Most folds have only one hinge between two
adjacent inflection points, although there are folds with many hinges (see below under "fold
style"). The relatively straight domains between inflection points and the hinge are referred to
as the FOLD LIMBS and the region of high curvature around the hinge is referred to as the
HINGE ZONE or the NOSE of the fold. The line passing through the hinge parallel to the axis
is the HINGE LINE.
Folds are referred to as
ANTIFORMS
if
they
close
upwards, SYNFORMS if they
close downwards, and NEUTRAL
FOLDS if they close sideways.
These terms should not be
Figure 71
confused with the terms "anticline"
and "syncline" (although they
commonly are). The latter terms may be used only if both: the surface being folded is
bedding or volcanic layering and the facing direction of the surface is known. Under these
conditions a SYNCLINE is a fold in which the layering faces inwards towards the fold nose
and an ANTICLINE is a fold in which the layering faces outwards from the fold nose. Either
might be a synform, an antiform or a neutral fold (Fig. 71).
(b) Many Surfaces

Hinge lines may be defined

on each adjacent surface in a fold
consisting of more than one
surface. The surface that contains
the hinge lines from all adjacent
surfaces is the HINGE SURFACE
(or AXIAL SURFACE) of the fold
(Fig. 72).

Figure 72

A fold with a planar axial surface is a PLANE fold. Non-plane folds may be treated by
considering domains within them that are themselves approximately planar.

(c) Attitudes of Folds

The attitude of a fold is fully specified only when both the attitude of the fold axis and
the axial surface are given. The following are general terms used to describe the attitudes of
folds:
(i) Based on the dip of the axial surface:

EPSC 203 (2012) p. 70

UPRIGHT fold (dip greater than 75o)
INCLINED fold (dip 15 - 75o)
RECUMBENT fold (dip less than 15o)
(ii) Based on the plunge of the fold axis:
VERTICAL fold (plunge greater than 75o)
PLUNGING fold (plunge 15 - 750)
HORIZONTAL fold (plunge less than 15o)
(The divisions given are only a guide)
Thus, in describing a fold one may use two terms. e.g., "inclined, plunging".
Recumbent folds are, of course, always horizontal, and vertical folds are always upright, since
the fold axis must lie in the axial surface.
A special term RECLINED is used to describe a fold that plunges down the dip of its
axial surface; i.e., in which the pitch of the fold axis on the axial surface is greater than about
75o.
(d) Fold Style
Several features contribute to the "style" of a fold:
(i) Interlimb angle. This is the angle between the two limbs of the fold, measured across the
axial surface. A fold in which the interlimb angle is zero (limbs parallel) is an ISOCLINAL fold.
Terms such as "open", "moderately open" and "closed" have been used with specific
meanings, but it is probably better to give the range of interlimb angles.
(ii) Change in layer thickness
around the fold. Layers in a fold
may keep a constant orthogonal
thickness around a fold, in which
case the folds are referred to as
PARALLEL folds, or they may vary
in orthogonal thickness, generally
becoming thicker in the noses than
Figure 73
on the limbs.
An idealized
example of the latter situation is
that in which adjacent folded
layers have identical (congruent) forms. Such folds are known as SIMILAR folds, and have
the additional property that the "thickness" measured parallel to the axial surface is constant
around the fold (Fig. 73).
Although folds are generally described as either "similar" or "parallel", perfect

EPSC 203 (2012) p. 71

examples of the idealized types are rare, particularly in the case of similar folds. Folds
approximating the idealized parallel type are common in terrains of low-grade metamorphism.
A more complete description of the changes in layer thickness around folds is provided by a
description of their "isogon" patterns.

An ISOGON is a line
joining points with the
same attitudes on
adjacent layers in the
profile of a fold (Fig.
74).
Figure 74. Isogon patterns for parallel and similar folds.
Thus, parallel folds have isogon patterns that are convergent towards the noses of the folds
("convergent" isogon patterns) and similar folds have parallel isogon patterns. It is readily
apparent that these are just two special cases. The full spectrum of possibilities involves five
classes of isogon pattern (Fig. 75).

Figure 75

EPSC 203 (2012) p. 72

Figure 75

The major control on the

constancy of layer thickness
around a fold is the competency
(resistance to flow) of the rock.
Thus, strong rocks maintain
broadly constant layer thicknesses,
producing parallel folds.
It is
because rocks are strong at low
temperatures that folds produced
Figure 76. a Isogon patterns in a fold in rock consisting at low temperatures are commonly
of alternating strong (resistant to flow) and weak layers. approximately parallel. At higher
b Chevron fold.
temperatures
rocks
become
weaker and tend to flow towards
the noses. However, at a given
temperature, some rocks are always stronger than others. In a layered sequence the
relatively strong rock will consequently display less of a tendency to flow into the noses than
the weaker rock, and the result is a rock that consists of alternations of layers with small
changes in thickness around folds, with weakly convergent isogon patterns, and layers
exhibiting marked thickening into the noses, with divergent isogon patterns. Most "similar"
folds are like this. To produce a true similar fold would require that there be no mechanical
difference whatsoever between adjacent layers. Although mechanical contrasts are
diminished at higher temperatures they are rarely completely eliminated (Fig. 76a).
(iii) Development and character of fold nose.
This characteristic gives rise to a number of special terms used to describe particular
types of fold:
CHEVRON FOLD: A fold with a very sharp nose and straight limbs. The term is commonly
restricted to folds that are essentially symmetrical, with interlimb angles around 60o.
Asymmetric folds of this style are referred to as KINK FOLDS. In both cases the folds require
the existence of a layered sequence in which relatively strong layers are separated by weak
inter-layer surfaces (Fig. 76b).

EPSC 203 (2012) p. 73

KINK BANDS are kink folds in which
the short limbs are repeated in a regular
fashion for large distances parallel to the
axial surface. The term is commonly
reserved for structures with wavelengths
less than about 20 cm (Fig. 77).

Figure 78. Ptygmatic folding.

Figure 77. Kink band.

PTYGMATIC FOLDS : Folds that demonstrate

extreme thickening in the noses and "negative"
interlimb angles. They resemble the loops of
an intestine. They are produced in isolated
layers of relatively competent rock that have
been constrained to shorten a great deal
because of the weakness of the rocks around

them (Fig. 78).

INTRAFOLIAL FOLDS: Rootless
tight folds preserved between
layers in a foliated rock. Such
folds are a result of extreme
deformation,
which
has
transposed the limbs of the folds
parallel to the foliation. Large
amounts of movement on the
limbs then leave the original
closures isolated (Fig. 79).

Figure 79. Intrafolial folds.

MULTIPLE HINGE FOLDS: folds in which

there is more than one curvature maximum
between inflexion points. Commonly there
are two and such a fold is referred to as a
BOX FOLD. They are formed in conditions
similar to those responsible for the formation
of kink folds and kink bands and are
commonly associated with them (Fig. 80).
Figure 80. Box fold.

EPSC 203 (2012) p. 74

(iv) Degree of cleavage development. Whether there is an axial-plane cleavage, whether it is
fanned, what type of cleavage it is and how intense it is all contribute to the "style" of a fold.

The importance of fold style is partly in the description of folds, but more importantly in
the correlation of folds of a particular folding event. Since it may reasonably be assumed that
rocks in a given area undergoing a given deformation event were under approximately the
same physical conditions, they may be expected to exhibit similar styles of folding. This is
because the style of the folding is ultimately controlled by the physical conditions under which
the deformation occurred. However there are some cautionary notes to bear in mind:
1. The style is governed as much by the lithologies involved as by the physical
conditions. Style may be compared only between folds in the same lithological type.
Furthermore, since the other lithologies with which it is associated also control its
behaviour (e.g., it may be the strongest or the weakest lithology in a given layered
sequence) it is important also to take account of the associated rocks.
2. The style is also governed by the intensity of the deformation, and this may vary
markedly through the area in a given event. This is a harder problem to avoid, and can
lead to a great deal of confusion in the correlation of folds of different events.
Notwithstanding these caveats, fold style is one of the most powerful methods by
which folds of a given event are correlated. Following a correlation, a systematic treatment of
the variation in attitudes of the folds over a region may then indicate whether or not the initial
correlation was sound. In practice, style and attitude are used in tandem in unravelling the
structural evolution of a region.

EPSC 203 (2012) p. 75

FOLD PROJECTION
We may wish to know how a fold, observed at the surface, would appear on a vertical
section at some depth. Alternatively, we may wish to provide a stylistically valid impression of
the fold, by viewing it in its profile plane. In order to do this we must project the fold, using the
information available at the surface, onto the surface on which we wish to study it.
The commonest, and most universally applicable, method is to project the fold down
its plunge (or up its plunge) onto the surface of interest. In essence, various points on the fold
are traced down the plunge until they intersect the target surface, and are then joined to
produce the fold on the target surface. The only assumption inherent in this procedure is that
the fold is cylindrical.
There are two simple ways in which to prepare a down-plunge projection:
(a) Grid Superposition
This method is the most rapid, but may be used only when there is no significant
topographic interference with the fold's appearance at the surface.

Consider a fold exposed on a flat

surface. The surface is in general not
perpendicular to the plunge of the fold.
We wish to project this fold onto a
vertical section with strike perpendicular
to the trend of the fold. In this projection
process, consider a horizontal line
perpendicular to the trend of the fold. It
may be projected down the plunge
without any distortion or length change
(Fig. 81).

Figure 81. Projection of a line AB, that is

perpendicular to the plunge line, in the downplunge direction. AB projects to A'B', where
AB=A'B'.

Now consider a horizontal line that is parallel to the trend of the fold. In the projection
process it remains straight, but is shortened (Fig. 82).

EPSC 203 (2012) p. 76

Figure 82. Projection of a line that is parallel to the fold trend, in the
down-plunge direction. AB projects to A'B', where A'B'=AB tan .
Consequently, if we impose a grid on the fold at the surface, with lattice lines parallel
and perpendicular to fold trend, we may redraw the grid according to the changes in length of
the grid elements as defined above. Since the transformation we have effected is
homogeneous, the fold should also suffer a similar homogeneous transformation. Points on
the fold that occupied a particular position in a particular grid square will project to points that
occupy analogous positions in the transformed grid square. In this way we may reconstruct
the appearance of the fold in the section (Fig. 83).

Figure 83. Down-plunge projection by gridding.

(b) Projection of individual points.

This method, although slower than using a grid, may be applied to folds in which the
surface topography has substantial effect on the appearance of the fold.

EPSC 203 (2012) p. 77

(i)

Without topography

The method consists of projecting individual points on the fold, parallel to fold plunge,
until they intersect the target surface. This is done on an orthographic projection of the
vertical section containing the fold-plunge, from which the distance down the target section
may be determined. The lateral position on the target section may be determined from the
plan view (Fig. 84).

Figure 84

EPSC 203 (2012) p. 78

This method may readily be
adjusted
for
topography,
since
topographic height may be included on
the orthographic projection (Fig. 85).

Figure 85

EPSC 203 (2012) p. 79

Projection in Directions Other Than Plunge
Such projections may be performed only under conditions in which a particular style
for the folding may be assumed, such as ideal similar folding or ideal parallel folding. Low
temperature deformation may produce folds that are nearly ideally parallel, but ideal similar
folds are rare. We shall concern ourselves only with the extrapolation of parallel folds, a
procedure that is commonly used in foreland-fold-and- thrust belts like that of the Canadian
Rockies. The commonest method for such reconstructions is that due to Busk, in which the
folds are built up from successive segments of concentric folds.
Consider
a
set
of
measurements, collected on a
traverse of the profile plane of a
fold in an area of parallel folding.
The folding may be reconstructed
from these measurements by
considering the folds to be made
up of a series of circular arcs
passing through all the measured
data.
These arcs may be
constructed by identifying their
centres
as
the
points
of
intersection of lines passing
Figure 86
through each of two adjacent
measurements, perpendicular to the traces of the bedding at each point (Fig. 86).
In this way a complete set of circular arcs may be constructed and, assuming parallel
folding, similar arcs, constructed with the same centres, may be added at greater depth.

Extrapolation to depth ultimately

gives rise to problems in the form of
geologically
unreasonable
cuspate
structures (Fig. 87).
Figure 87

EPSC 203 (2012) p. 80

In the practical
situation
these
structures
are
compensated for either
by faulting in the noses
of the folds, or by
disharmonic
folding
(Fig. 88).

Figure 88
In constructing such compensating structures in cross sections, the principle that is
generally used is that the lengths of adjacent beds between locations of zero slip (such as the
noses of open folds) should be constant. Cross sections adjusted in this way are "balanced".
The method described above for the reconstruction of folds makes no allowance for
the possibility that the beds whose attitudes were measured may have been at different
stratigraphic positions. That is, all measurements were joined to produce a structure, without
considering whether they had been collected on the same bed. In circumstances in which the
same bed is identifiable at two different places on the traverse, a dip may be interpolated
between two adjacent measurements, in order to force the fold to join points at which the
same bed was observed: Consider two adjacent bedding measurements, at A and B, where
the dip is greater at B than at A (Fig. 89).
Construct the normals to the dips
at each, find their intersection (C),
and extend the normals beyond it.
Construct
the
perpendicular
bisector of AB, to intersect AC at Z
Place a point O on the normal AC,
beyond Z
Find D on BC, such that BD = AO
Find P, at the intersection of the
perpendicular bisector of line DO
with BD
Figure 89

EPSC 203 (2012) p. 81

O and P are the centres of two tangent arcs that define the curve required to force the fold
through both A and B.
Another method that has been used
to reconstruct parallel folds is due to Suppe.
This method, instead of building up folds as
additions of adjacent circular arcs, considers
them to consist of long straight segments
separated by kinks. The styles of the folds
so produced conform more closely to those
observed in many foreland fold-and-thrust
belts.
In this method the profile plane is
divided into regions of approximately
constant dip, by inspection.
Adjacent
domains of different dip are considered to be
Figure 90
related by a kink, the axial surface for which
bisects the angle between the dips of the
two domains. The assumption of symmetry
in the kink is not critical, and may be relaxed if stratigraphic information requires it. Where two
axial surfaces defined in this way intersect, the data from the middle domain are ignored, and
a new axial surface attitude is defined from the two outer domains. In this way a complete
section at depth may be constructed. Because the method produces kink-folds it does not
produce the same geometrical problems when extrapolated to depth - kink-folds may be
extrapolated an unlimited distance. However, balancing the section may still require faulting
within the structure (Fig. 90).

EPSC 203 (2012) p. 82

CLEAVAGE
Cleavage - the tendency of a rock to break along closely spaced planes - is a type of
foliation produced in rocks during strain. The distinction is commonly made between
"penetrative" and "non-penetrative" cleavage, depending upon whether domains that are free
of the cleavage are present. This distinction is dependent on the scale of the observation;
there are few cleavages that are truly penetrative on a microscopic scale. In practice, when
unqualified, the term "penetrative" refers to "naked-eye" observations on a hand specimen.
Penetrative cleavages are referred to as "slaty cleavages" when the grain-size is too
fine to be seen, and "schistosities" when the grains responsible for the cleavage are visible.
Non-penetrative cleavages are referred to as "spaced cleavages" or "fracture cleavages" (the
former term is preferred).
The type of cleavage developed in a rock is a function of the rock type, the amount of
strain experienced, and the (grain-scale) mechanisms by which the strain was achieved.
(a) Rock Type
In general, the best cleavages are developed in rocks that contain significant
proportions of minerals that have high aspect ratios (long-dimension to short-dimension
ratios). The common mineral species with high aspect ratios are micas (plates) and
amphiboles (needles or prisms). Thus, mica-rich metasediments, which are generally derived
from fine-grained (clay-rich) sediments, have well-developed cleavage, whereas
metasandstones, which do not generally contain large amounts of mica, do not, even if they
are deformed under the same conditions and have experienced the same amount of strain. It
is common for slaty cleavage in deformed fine-grained sediments to pass into spaced
cleavage in adjacent sandstones.
(b) Amount of Strain & Mechanism
Cleavage is for the most part a reflection of the preferred orientation of the long axes
of grains with high aspect ratios, although it is in some cases enhanced by compositional
banding as well. This preferred orientation may be brought about by three different
mechanisms, operating independently or in concert:
(i) Bulk rotation of high-aspect-ratio grains during strain
Homogeneous strain produces the rotation of almost all lines in a body towards the
plane of flattening. Thus, a homogeneous strain could effect a preferred orientation of
platy minerals on the plane of flattening, or of the long axes of needles on the plane of
flattening, simply by rotation during strain. It is possible to show that this should occur
whether the grains act as simply passive markers in the rock, or as more rigid bodies
moving in a weaker medium. (The latter model probably corresponds more closely to
the true situation, at least for mica-rich rocks, since the matrix quartz grains will

EPSC 203 (2012) p. 83

generally deform more easily than the micas.) Some well-developed slaty cleavages
may be developed entirely by this mechanism.
(ii) Strain-induced recrystallization
During strain, grains that are unfavourably oriented with respect to the bulk strain axes
(for example micas whose basal planes are at a high angle to the flattening plane)
may experience greater internal strain (of their lattices) making them favourable grains
for degradation by transfer of their components along grain boundaries towards more
favourably oriented grains, resulting in a net increase in the proportion of favourably
oriented grains. Alternatively, grains that are kinked because of their unfavourable
orientations may recrystallize into two grains (the two limbs of the kink) each of which
has an orientation more favorable to that of the bulk strain.

(iii) Stress-induced neocrystallization

If new mineral grains are growing during the deformation ('neocrystallization', as
distinct from the recrystallization of existing grains) only those grains favourably
oriented with respect to the stress field will tend to grow, resulting in the development
of new grains all of which have a well developed preferred orientation.
Criteria by which the relative importance of these three mechanisms may be
determined are difficult to devise. It is probable that all are important to varying degrees in
different conditions. For example, at low temperatures, where crystallization is slow, (i) may
be paramount, whereas a rock deformed at high temperatures under conditions far removed
from those at which the mineral assemblage of the rock is in equilibrium may be dominated by
(iii).
Mechanisms (i) and (ii) will tend to produce cleavages that are parallel to the plane of
flattening. Mechanism (iii) will produce cleavages dictated by the stress field, which may not
exactly correspond to the strain field (if the rock is anisotropic). However, the differences may
be small, and may be reduced if the existing cleavage is rotated towards the plane of
flattening by continuing strain. Thus, the assumption is usually that the cleavage surfaces are
approximately parallel to the surface of flattening for the strain. Note, however, that if the
strain is rotational the cleavage may reflect only the attitude of the plane of flattening for the
last part of the strain, rather than for the total strain, and there are also other possibilities,
depending on the rate at which the cleavage can keep up with the rotation of the plane of
flattening. Since the strain in fold limbs is generally rotational, small departures of the
cleavage plane from the plane of flattening may be expected, and even overprinting of two
cleavages of slightly different attitudes in some cases.
For all three mechanisms the quality of the cleavage should reflect the degree of
strain, at least until a level of strain is reached at which the cleavage is essentially perfect.

EPSC 203 (2012) p. 84

Cleavage defined by compositional domains
Although almost all cleavage reflects a preferred orientation, it is common for cleavage
to be accompanied by oriented domains of different composition. For a spaced cleavage this
is reflected in the concentration of micas, for example, in the planes that define the cleavage.
Even a slaty cleavage may exhibit such domains on a microscopic scale.
The best developed compositional
banding is associated with crenulations, in
'crenulation cleavages' (Fig. 91).
Crenulation cleavages are developed
from the limbs of microfolds ('crenulations')
Figure 91. Crenulation cleavage
in
pre-existing,
usually
penetrative,
cleavages. Development of the microfolds is
commonly accompanied by migration of quartzofeldspathic material out of the limb regions,
which are experiencing greater stress, leading to an apparent enrichment of the limb-regions
in micas. In extreme cases the limb regions may become fractures, entirely filled with micas,
separating nose- (or long-limb-) regions much poorer in micas. This compositional banding
induced by metamorphic differentiation may be mistaken for bedding-inherited banding in
rocks that subsequently experience a high-grade recrystallization which conceals the
crenulation-related origin.
Crenulation cleavages are very common in regions that have been deformed at
relatively low temperature after a major phase of deformation in which a penetrative cleavage
was produced.

EPSC 203 (2012) p. 85

LINEATIONS
Lineations are simply linear features in a rock. They may be of primary origin, such as
the preferred orientation of the long axes of detrital quartz grains parallel to a current direction
in sediments, or of tectonic origin.
Intersection Lineations
Intersection lineations are lineations produced by the intersection of two planar
features, such as bedding and cleavage. They are readily observable and measurable in the
field, since they may be seen as, for example, the trace of bedding on a cleavage surface or
the trace of cleavage on a bedding surface.
Mineral Lineations
Mineral lineations are lineations reflecting the preferred orientation of the long axes of
mineral grains or of aggregates of mineral grains with high aspect ratios. They may be
produced by any or all of the mechanisms discussed above for the development of cleavage
and are consequently commonly parallel or nearly parallel to the long axes of the strain
ellipsoid.

EPSC 203 (2012) p. 86

CLEAVAGE AND FOLDING
The growth of folds is a response to layer-parallel shortening of competent layers in an
incompetent medium. In the general case, in which the principal compressional stress
responsible for this shortening does not lie in the plane of the layering, the fold axis develops
along or close to the line of intersection of the plane of flattening and the plane of the layering.
The fold will generally grow initially orthogonal to the layering, so that its axial surface will not
develop parallel to the plane of flattening except in special cases. However, once the fold
develops significant amplitude its axial surface will rotate towards the plane of flattening. In
general, then, the axial surface is approximately parallel to the plane of flattening for the bulk
strain. Given the considerations above, it may therefore be expected that the cleavage will
also parallel the axial plane. This relationship is commonly observed, and such a cleavage is
an 'axial plane cleavage'.

Figure 92. Cleavage fans

Significant departures from perfect parallelism of cleavage with the axial planes of
folds are common, however. Cleavages are commonly fanned about the fold axis, with
attitudes to either side of that of the fold axis on adjacent limbs of the fold. In the hinge
surface they are parallel to the axial surface (Fig. 92).
The fanning of cleavage is a reflection of local strain distributions in the rock in which
the planes of flattening are not parallel to that of the bulk strain. These local distributions arise
because of competence contrasts between adjacent layers. Competent layers tend to
experience strain whose axes are locally at high angles to the layer boundaries, thereby
developing convergent cleavage fans. Adjacent incompetent layers, which experience large
amounts of shear imposed by their neighbours, exhibit divergent fans (Fig. 93).
The change in cleavage attitude as a layer boundary is crossed is known as 'cleavage
refraction'. In the absence of bedding data, the attitude of cleavage in an area may provide
an indication of bedding attitude provided the type and attitude of the fan for the lithology in

EPSC 203 (2012) p. 87

which the cleavage is observed is known.
Since the cleavage, even when it is fanned, contains the fold axis, the line of
intersection between bedding and cleavage is parallel to the fold axis (all bedding contains the
fold axis). This intersection lineation between the bedding and the cleavage is therefore a
simple and accurate measure of the attitude of the fold axis in a cylindrically folded terrain and
is very widely used.
Note that it may be used even if the folds
themselves are never observed. Likewise,
the attitude of cleavage may be used to
provide an indication of the attitude of the
axial surface. If the cleavage is fanned, its
attitude may depart from that of the axial
surface by as much as 30o in the limbs of
folds, although in the hinge-surface region it
will always be parallel to it. Equation of the
attitude of the cleavage with the axialsurface attitude must therefore be done with
care. A more direct and reliable measure of
the axial surface attitude is provided by
measuring the trace of the axial surface on
an exposure in which the fold can be seen,
Figure 93. Cleavage refraction in folds
and determining the attitude of the plane that
contains this trace and the fold axis (Fig. 94).

Figure 94. Use of axial trace and fold axis to determine the attitude of the axial surface. On
the stereogram, B is the fold axis.

EPSC 203 (2012) p. 88

EPSC 203 (2012) p. 89

FOLDS ON THE STEREOGRAM
Beta Plot
If the attitudes of two limbs of a fold are known, the intersection of their great-circle
traces on a stereogram is the fold axis. A diagram of this kind, in which fold limbs are drawn
as great circles, is known as a "beta plot", and the corresponding fold axis is 'beta' (Fig. 95).

Figure 95
Pi Plot

If many data are available for the

fold, the intersections of the several great
circles give rise to a confusing number of
points. The data may more simply be
plotted as poles to the planes. In this case,
since every plane contains the fold axis, the
pole of every plane must be perpendicular to
the fold axis, and the poles trace out a great
circle - the plane perpendicular to the fold
axis. This plane, the 'profile plane', may be
identified, and its pole is the fold axis. A plot
of this kind is known as a 'pi plot' (Fig. 96).

Figure 96

EPSC 203 (2012) p. 90

Interlimb Angle
If bedding data were collected completely randomly around a fold and plotted on a pi
diagram the poles would be distributed on the profile plane in two clusters, reflecting the two
relatively straight limbs, and a broad rarefied distribution of points representing bedding data
measured around the hinge of the fold.
The region between the two limbs, where
the axial surface lies, would be reflected in
an unoccupied region on the profile plane.
The interlimb angle for the fold may be
measured either between the poles to the
two limb clusters across the unoccupied
region, or between the intersections with the
profile plane of the two great circles
representing the fold limbs (across the great
circle representing the axial surface) (Fig.
97).

Figure 97

Cleavage in a fold occupies

the region not occupied by the
bedding, and will be distributed on
the profile plane if it is fanned, the
extent of the distribution reflecting
the extent of the fanning. For a
divergent fan the cleavage in a
given limb is displaced from the
axial surface attitude towards the
limb in which it is measured. For a
convergent fan the opposite is true
(Fig. 98).

Figure 98

EPSC 203 (2012) p. 91

Lineations in Folds
Since fold axes commonly develop perpendicular to the shortening line in the layering
(i.e., close to the plane of flattening for the bulk strain) there is generally extension parallel to
the fold axis in the layering. Thus, a mineral lineation may develop in the direction of the fold
axis. This is the commonest relationship observed, and it provides a useful estimate of the
attitudes of fold axes in intensely deformed terrains. There are, however, other possibilities.
In particular, it is possible that local strain within the fold may develop a stretching direction
that is at a high angle to the fold axis (for example due to shearing of incompetent material on
each limb). In cases such as this, the lineations would generally be expected to vary in
attitude around the fold, so that a careful examination of the structural data should reveal the
relationship.
Intersection lineations, as previously discussed, should in general be parallel to the
fold axis.
Two other kinds of mesoscopic lineation are common in association with folds, and
both are generally parallel to the fold axis:
Rods
Rods are cylindrical bodies of a material other than that making up the rock as a
whole, e.g. vein quartz. They may develop during folding due to the opening of cavities in the
noses of the folds between adjacent competent beds, or within competent layers at the noses
due to extensional failure there. The cavities are then filled with material deposited from
intergranular fluids.
Mullions
Mullions are cylindrical bodies of
material of the same composition as the rock
as a whole, bounded by surfaces of some
kind. Cleavage mullions are prismatic, and
bounded by a bedding-parallel jointing and
an axial-plane cleavage. Fold mullions are
minor folds with distinctly different forms on
their convex and concave surfaces - very
broad on the surface convex towards a
relatively incompetent layer and very sharp
on the other. They are commonly seen as Figure 99 a Cleavage Mullion. b Fold Mullion
ornaments on the surface of the competent
layer, and are due to disharmonic buckling during the folding process (Fig. 99).

EPSC 203 (2012) p. 92

FOLD MECHANISMS
One must distinguish between idealized fold mechanisms, and the actual mechanisms
by which most folds grow.
IDEALIZED MECHANISMS
(a) Flexural folding
Flexural folding is folding in which the directions of movement within the rock are
parallel to the layer being folded (Fig. 100).

Flexural folding may be considered

as "flexural-slip" folding if the majority of the
movement takes place on preferred planes
of weakness parallel to the layering. In
"flexural-flow" folding, on the other hand,
there are no discrete planes of weakness,
and the deformation is distributed throughout
the rock.
In flexural-slip folding the nature of
the (relatively small) strain that occurs
between the planes of weakness is not
specified. The simplest type of such strain
would be that in which principal strain axes
were everywhere parallel or orthogonal to
the layering. Such a strain distribution,
known as "tangential longitudinal strain",
illustrates some typical properties of flexuralslip folding (Fig. 101):

Figure 100

(a) stretching on the outer part (extrados) of

the fold
(b) flattening on the inner parts (intrados)
(c) the existence of a neutral surface, on
which there is no deformation.
Figure 101

Coherently deformed domains within

flexural-slip folds commonly exhibit an
approximation to these conditions, although

EPSC 203 (2012) p. 93

modifications to the tangential longitudinal strain condition, in particular in the form of extra
flattening onto the axial surface, are common.
The isogons of a fold that has experienced tangential longitudinal strain are
convergent, like those of a parallel fold (Class IB).
Flexural-slip folding leads to the redistribution of any pre-existing lineations. For the
common case, in which the preexisting lineation lies in the layer being folded, a lineation that
was present within a neutral surface in the coherent domain of the rock would experience a
simple rigid-body rotation, maintaining a constant angular relationship to the fold axis. Thus, it
would be distributed on part of a small circle about the fold axis (filled circles, Fig, 102). On
the extrados of the fold, however, because of the stretching perpendicular to the fold axis, the
angle between the fold axis and the lineation increases in the hinge zone, and the distribution
is not precisely that of a small circle (see dashed line, Fig. 102). By the same token, on the
intrados of the fold, the angle between the lineation and the fold axis decreases in the hinge
zone.
Note that in no case is the entire
small circle occupied.
Flexural-flow folding is essentially the
limiting case of flexural-slip folding, in which
the domains exhibiting coherent deformation
have become vanishingly thin.
In
consequence, the features of the extrados
and the intrados of a coherent domain
become very similar to each other, and
deformation within the fold is like that along
a neutral surface, with maintenance of a
constant angular relationship between the
lineation and the fold axis.
(b) Shear folding
Shear folding, also known as
"passive folding" is folding in which the
Figure 102
movement takes place at some angle to the
plane being folded, in the axial plane of the fold. For planes of slip that are sufficiently closely
spaced this gives rise to a smooth fold (Fig. 103).

EPSC 203 (2012) p. 94

Such folding should give rise to
isogon patterns that are parallel (Class II),
and to perfect similar folds. Indeed, it is
because similar folds are much in evidence
in the field, and because of the widespread
evidence of slip on axial-plane cleavages
that such folding has been thought to be
important.
In addition, shear folding should
redistribute a pre-existing lineation on a
great circle, containing the original lineation
direction and the vector of slip.
Figure 103
REAL FOLD MECHANISMS
Real folds generally have isogon patterns of Class III (divergent) or Class IC (less
convergent than true parallel folds). In addition, great-circle distributions of lineations due to
refolding are rarely observed except in ductile shear zones. It would appear therefore that
shear folding is not an important mechanism in general. Most folding is now thought to be
flexural, at least in its initial stages.
Class IC isogons can be produced by adding a component of homogeneous flattening
onto the axial plane to the simple flexural models given above. Thus, one may view folding in
rocks consisting of layers of alternating competence as being of the flexural-slip type, in which
the stronger layers are equivalent to the coherent domains of the flexural slip model, but have
experienced additional flattening onto the axial surface. The weaker layers, on the other
hand, are equivalent to the preferred planes of weakness in the flexural-slip model. Their
Class III isogons are constrained by the geometry of the layers around them.

FOLD MECHANISMS - SOME GENERALIZATIONS FROM MODELLING

The following generalizations are based on the experimental buckling of strong layers
in a weaker matrix, and on mathematical modelling.
(a) Buckling occurs only after a certain amount of homogeneous shortening. There is more
shortening for lower viscosity contrasts.
(b) There is a dominant wavelength for buckling, dependent on the viscosity and the thickness
of the layer.

EPSC 203 (2012) p. 95

1

1 3
= 2t

6 2

large thickness produces high wavelength

high viscosity contrast produces long wavelength

(c) Additional homogeneous strain is developed during and after buckling.

EPSC 203 (2012) p. 96

JOINTING
Mechanics
Joints were defined earlier as discrete fractures across which there has been little or
no movement. They are distinct from (brittle) faults only in that the amount of movement is
small. They reflect failure of a rock under a stress field that is sufficient to break the rock, but
which may then be dissipated by a small amount of movement on the joint surface. In
general, this dissipation of the stress field is achieved because the stress field is a result of
internal stresses developed in the rock, rather than external ones. Faults result from failure
of a rock under an externally applied stress field.
Failure of a rock in a stress field may occur in one of two ways: (i) perpendicular to the
least compressional stress (i.e., in an effectively extensional environment) in which case the
fracture will open perpendicular to its walls, and (ii) on planes on which there is a high
resolved shear stress (usually planes containing the intermediate principal stress direction) in
which case there is slippage parallel to the walls. This gives rise to two types of joint,
classified on a genetic basis: "extensional" and "shear" joints. Unless minor displacement is
visible across joints, they may not be classified in this way without recourse to some indirect
reasoning.
Whether
an
extensional or a shear joint
will be produced in a given
environment depends on the
size of the deviatoric stress
relative to the Mohr envelope
for failure. A small deviatoric
stress will give rise to
extensional joints whereas a
larger one will give rise to
shear joints (Fig. 104).
Many
joints
are
produced in rocks as a result
of the extensional stresses Figure 104. Failure conditions for joints. Circle A reflects the
developed in them as they stress conditions for conjugate joints a dihedral angle apart.
cool. In the case of igneous Circle B illustrates the stress condition leading to extensional
rocks, this cooling is simply a failure.
result of the long-term
dissipation of magmatic heat. In the case of metamorphic rocks, it may reflect cooling during
the erosional exhumation that brings the rocks to the surface. In general, because of the
relatively low thermal expansibility of rocks, deviatoric stresses induced by cooling are small,
so that they do not give rise to shear joints. They will give rise to extensional joints when the
rocks have been unroofed sufficiently for the extensional strength of the rock to be overcome.

EPSC 203 (2012) p. 97

At that stage, extensional failure will relax the internal stress in one direction. Continued
cooling or exhumation will then allow stress release in another direction, commonly
orthogonal to the first. Such processes give rise to the commonly observed orthogonal joint
sets in granitic terrains.
Shear joints reflect the response of rocks to stored elastic deviatoric stresses of
greater magnitude than those that produce extensional joints. These stresses might be
produced by a small deformation event at depth, that gave rise to no ductile deformation, or to
stresses developed at shallow crustal levels, or they may be residues of the stress fields that
gave rise to observable ductile deformation such as folding.
In the latter case,
the stored elastic stress is
simply the equivalent of the
"elastic" strain and the
"primary" creep discussed
in connection with rheology.
These elastic strains are
equivalent
to
internal
stresses. At the depth at
which ductile flow occurred
the Mohr envelope was
narrow (Fig. 105), and its
width is a measure of the
magnitude of the stored
elastic stress.
With
Figure 105
unroofing,
the
Mohr
envelope widens.
The
elastic stress remains as a deviatoric stress (represented by the circle on Fig. 105). With
unroofing, as the confining pressure is released the Mohr circle migrates to lower stress, and
ultimately intersects the Mohr envelope at shallow depth, giving rise to (commonly conjugate)
shear joints.
There may, then, be simple relationships between the orientations of joint sets and the
orientations of stress systems that produced folding, although the folds and the joints may
have been produced at very different depths and times. If shear joints occur in conjugate
sets, their orientations allow estimation of the configuration of the stress field, since the
conjugate sets both contain the intermediate stress axis, and the smaller dihedral angle
between them spans the greatest compressional stress axis, as in the case of conjugate fault
sets

EPSC 203 (2012) p. 98

Shear joints may be classified
according to their relationships to fold
geometry, as "longitudinal" (Fig. 106) or
"cross" joints (Fig. 107). The differences
between these two configurations of joint
sets reflect the different possible orientations
of the stress field relative to the fold.
Extensional
joints
are
quite
commonly developed in folded terrains
perpendicular to the fold axes. Such joints
are known as "a-c" joints after an old
geometric terminology for folds in which the
fold axis is "b" and the other two orthogonal
axes are "a" and "c".

Figure 106. Longitudinal joints

Because joints reflect stored elastic

stress and the response of rock to this
stress, in a homogeneous rock they are
commonly uniformly distributed, and have
different spacings for different rock types.
Because motion on faults may disturb a
pre-existing stored stress, by imposing a
new one, joint concentrations and attitudes
in given rock types may change markedly as
faults are approached. This provides a
powerful means for inferring faults in poorly
exposed terrains.

Figure 107. Cross joints

EPSC 203 (2012) p. 99

CLEAVAGE & LINEATION ON A STEREOGRAM

CLEAVAGE
Being a planar feature, cleavage can be treated in much the same way as bedding.
Two apparent dips are combined on a stereogram, to give the attitude of the cleavage plane.

LINEATION
Lineations are observed in a rock only if the surface of observation is at a relatively
shallow angle to the attitude of the lineation. i.e., sections through a rock at a high angle to
the lineation will reveal no apparent fabric. This feature alone is sufficient to distinguish
lineations from foliations.

L
L

Figure 108
The attitude of a lineation may be determined if the lineation is observed in two
differently oriented sections through the rock. This is because the lineation lies in the plane
that contains:
(a) the apparent lineation L, as observed in the surface.
(b) the normal to the surface of observation, because L is the projection of L onto the surface
(Fig. 108).
Thus, the attitude of the lineation may be determined from a beta plot, if two differently
oriented outcrop surfaces, each carrying an apparent lineation, are observed. If several data
sets are available, an equivalent pi plot may be prepared, in which the poles to (n-L) planes
are plotted, rather than the planes themselves.

EPSC 203 (2012) p. 100

CONTOURING ON A STEREOGRAM
When large amounts (>30) of structural data are available they are commonly
presented not as a pi plot, but as a "contoured" plot, in which the distribution of pi poles has
been contoured on an equal-area net.
In order to be able to contour a net, a value must be assigned to every grid point on
the net. These values are assigned by counting the number of data points within a specified
distance of the grid point. Commonly, the specified distance is chosen so that the area being
sampled around the grid point is equal to 1% of the total area of the net. Clearly this method
would have obscure significance if a Wulff net were used. Contouring is always done on an
equal area net. Once a number has been assigned to each grid point the net may be
contoured, using contour intervals of e.g., 5 data points or 5% of total data points.

The size of grid used, i.e., the number of

grid points, determines only the smoothness of
the contour lines. Normally it is chosen so that
a grid square is significantly smaller than the
sample area. The only limitation on grid points
is the time available for contouring.
Contouring may be done by hand, by
moving a template with a circular hole in it the
size of the sample area over each grid point
and counting the data points visible within the
hole (Fig. 109a), but this is time-consuming and
cumbersome. Most contouring is done by
computers.
A line with very shallow plunge, although
it plots near to the primitive on only one side of
the stereogram, should be considered as a
member of the data set on the diametrically
opposed side of the stereogram. In computer
programs this is ensured by adding an extra
ring of data around the outside of the primitive
before assigning counts to the grid points.
When contouring is done by hand it is ensured
by using a special template, consisting of two
counting circles separated by a distance equal
to the diameter of the stereogram (Fig. 109b).

Figure 109

EPSC 203 (2012) p. 101

DRILL HOLE PROBLEMS

This section concerns the representation of possible attitudes for a planar feature
intersected by a drill, on a stereonet.
Given the attitude of a drill hole, and
the angle between the drill hole and the
planar feature intersected (which can be
determined from examination of the drill
core) the planar feature is known to have
one of several possible attitudes, all of which
are tangential to a cone about the drill
direction. The verticial angle of the cone is Figure 110. Possible attitudes for planes,
the angle between the feature and the drill tangent to the cone; from drill-hole data.
hole (Fig. 110).
If the angle between the drill hole and the planar feature is , then the angle between
the drill hole and the pole to the planar feature is (90-), and all possible poles to the feature
also lie on a cone centred about the drill hole. This cone has the verticial angle (90-). It is
simpler to monitor possible attitudes of the feature using the poles rather than the planar
attitudes. These poles may be represented on a stereogram as a small circle about the axis
of the drill hole. The small circle is the locus of all lines that are a constant angular distance
from the drill hole.

PLOTTING THE SMALL CIRCLE

(i) Compass-and-Wulff-net method
Using the fact that small circles plot as true circles on the true stereographic or Wulff
net, the small circle can be constructed rapidly with a compass. This is by far the quickest
method if (90-) is much less than 900 and also less than the plunge of the drill hole. Under
these circumstances the entire small circle lies within the primitive.

EPSC 203 (2012) p. 102

Procedure (Fig. 111):
(a)

Plot the drill hole.

(b)

Determine the attitudes of the

two extremities of the circle,
by measuring off (90-)
inwards and outwards from
the drill hole along the vertical
section containing the drill
hole.

(c)

Draw the circle whose centre

lies in the vertical section, and
which passes through both
extremities.

Note
that
the
measurement
performed on the stereogram gives DA < DB
so that D is not the centre of the circle on the
stereogram, although it represents the
geometric centre of the cone.

Figure 111

Figure 112. Perspective diagram of the two small circles for a cone, in stereographic
projection.
If the plunge of the drill hole is shallow, or if (90-) is greater than the plunge of the drill
hole, then the complete small circle does not lie within the primitive. It must therefore be
represented as two separate circular arcs, each lying within the primitive (Fig. 112). These
two circular arcs may still be constructed using a compass, although the method may become
quite cumbersome. In general, the second method would be preferred in this case.
AB and small circles 1 and 2 lie on the horizontal plane. So does the primitive. Thin
lines are those used for construction of the stereographic projection, passing through the top

EPSC 203 (2012) p. 103

pole of the 3D sphere. The thin-line segment OPQ on circle 1, which lies outside the
primitive, is replaced by the thick-line segment RST of circle 2.

Using
the
compass- and - Wulffnet
method
the
position of the second
segment must be
determined using an
accurate
(generally
scaled) diagram of
the vertical plane
containing the drill
hole (Fig. 113).

Figure 113

Figure 114

The resulting complete locus has one of three forms (Fig. 114).
(ii) Stepwise construction

EPSC 203 (2012) p. 104

Using the fact that all lines on

the cone are a constant angular
distance from the centre of the cone
(the drill hole) the small circle may be
constructed from a series of points
representing such lines on the
stereogram (Fig. 115a).

Procedure:

Figure 115

(a) Choose a succession of great circles passing through the drill hole.
(b) On each great circle, count out the distance (90-) in both directions from
the drill hole and mark the points.
(c) Join all points to produce a small circle.

This method may be used on either the Wulff or Equal-area net. In the latter case the
small circle produced is slightly distorted, especially if it falls near the primitive, although it is
still smooth. This method has great advantages in cases in which the small circle intersects
the primitive, because the location of the second segment may be determined directly using
the same method (Fig. 115b).

EPSC 203 (2012) p. 105

POLYDEFORMATION
REFOLDING A LINEATION
This generally gives something approximating a small-circle distribution, but distorted
by stretching or flattening during flexing, and by flattening onto the axial plane. Note that if the
new fold axis is parallel to the old lineation the latter is not redistributed. Note also that if the
two are perpendicular the small circle has a verticial angle of 900 and is therefore a great
circle.
REFOLDING A FOLIATION/BEDDING
We may treat a foliation like a lineation if, instead of monitoring the plane, we monitor
the pole to it. Thus, cylindrical folding is the special case in which folding of a pole distributes
the pole on a great circle. It is special because the fold axis is 900 from the pole, i.e., the fold
axis lies in the plane being folded. In the more general case, in which the fold axis does not
lie in the plane being folded (the fold axis is at less than 900 to the pole to the plane) the pole
to the plane is distributed on a small circle. This results in a conical fold in the plane, rather
than a cylindrical one.
In practice, a fold axis is unlikely to lie in a preexisting foliation unless that foliation is
mechanically active during deformation, so that any inactive surface will most probably be
conically, not cylindrically folded in a new deformation event. The converse is, however, not
true. Just because a surface is active during a deformation event it does not follow that it will
be cylindrically folded, i.e., it may still be folded about some axis that does not lie within it.
In analyzing complexly deformed terrains it is commonly necessary to predict the form
which will be taken by one structural element following rotation to various degrees about an
axis, or, conversely, to determine the attitude that element would have if the effect of a
rotation were removed. It is important to bear in mind that most deformation involves more
than simple, rigid-body rotations about axes. In particular, deformation commonly includes a
component of homogeneous strain. Nonetheless, rigid-body rotation is an important
component of much folding and the ability to perform rotations is an essential part of structural
analysis.
Rigid-body rotation is simply bulk rotation of a body about an axis. The term
"rigid-body" refers to the lack of internal deformation of the body, accompanying the rotation,
so that angular relationships that existed within a body before the rotation persist after it. This
is clearly an oversimplification for many deformation events, but, in the absence of an explicit
model for the nature of internal deformation it is the only model possible.
Fabric elements to be rotated may be planes or lines. When monitoring planes it is
possible, as usual, to monitor the pole to the plane, rather than the plane, in which case the
plane may be treated as a line. We need therefore concern ourselves only with the rotation of
lines. For a general rotation then, we have a rotational axis, a line that must be rotated about

EPSC 203 (2012) p. 106

that axis, the degree of rotation, and the direction of rotation (clockwise or anticlockwise
viewed down the plunge of the axis).
Many different techniques exist for performing rotations on stereograms. For simplicity
we shall treat only one here (that which is, in
my opinion, easiest to visualize).
This
technique relies on two properties of a
rigid-body rotation:
(a) Consider the plane that contains both the
rotational axis and the line to be rotated. For
a rotation of angle , the new position of the
line lies in this plane after it has been rotated
through the angle .
(b) Within this plane, the angle between the
axis and the line after the rotation is the
same as before the rotation (Fig. 116).

Thus, a rotation may be performed in

the following steps:

Figure 116

(a) Plot the rotational axis and the line/pole-to-plane that is to be rotated.
(b) Draw the plane that is common to them, and measure, in this plane, the angle between
the axis and the line (Angle on Fig. 117).

EPSC 203 (2012) p. 107

(c) Rotate the plane through the angle
about the axis, in the correct direction. Note
that this is simply done: since the axis lies in
the plane, the new position of the plane must
also contain the axis. This determines the
attitude of one line on the plane in its new
position. Another line on the plane in its new
position may be obtained in the plane
perpendicular to the axis, which we shall call
the normal plane. The trace of the plane on
this normal plane, T before deformation, is
rotated through angle on the normal plane
(i.e., the angle is calibrated on the normal
plane) by the rotation, to become Q. Thus,
Q and the rotational axis supply the new
position of the plane.

Figure 117

(d) In the new position of the plane, measure

off the angle from the rotational axis, to
determine the new position of the line.

Note 1: Any rigid-body rotation rotates a line on a cone. It therefore appears as a small circle
about the axis, of angle .
Note 2: If the rotation is such that the plane passes through the horizontal, the new position of
the line is on the small circle arc on the opposite side of the normal plane from its original
position (Fig. 118).

EPSC 203 (2012) p. 108

POTENTIAL FOR CONICAL FOLDS
Less work is generally required to
fold
a
mechanically-active
surface
cylindrically than to fold it conically. In
consequence, it is normal for first phase
folding of an active surface to produce
cylindrical folds. Passive planar features
folded in the same event will be conically
folded.
Second-phase
folding
of
a
mechanically-active surface may produce
either conical or cylindrical folds, or both.
Given the energy demands, cylindrical
folding is favoured. However, after the first
phase of folding there are at least two
attitudes for the bedding. If there is strong
mechanical coupling between the two F1
limbs (no weak material in the nose) then it
Figure 118
is probable that only one of the F1 limbs will
be cylindrically folded. The other limb is constrained to fold about the same axis as the first
limb which, since it does not lie in the second limb, produces conical folds. In this case, then,
we get cylindrical folds on one limb and conical folds on the other, both with the same axis.
If there is little or no mechanical coupling between the two limbs of the F1 fold, they
may deform independently. In this case each limb is likely to be cylindrically folded, but the
axes for the two new folds will be different, each lying at the intersection of the second axial
surface with the respective F1 limb. All transitions between these two situations are possible
(Fig. 119).

EPSC 203 (2012) p. 109

(a) Limbs 1 & 2 may fold independently,
cylindrically, about A and B.
(b) Limb 1 may fold cylindrically about A,
Limb 2 may fold conically about A.
(c) Both limbs may fold conically about C.

Figure 119
RECOGNITION AND DESCRIPTION
OF CONICAL FOLDS
Conical
folds
are seldom
recognized, for two principal reasons:
(a) Sections through them, as observed
in the field, are indistinguishable from
those of cylindrical folds.
(b) The full small circle for the poles to
the bedding is never traced out. Under
conditions of extreme deformation half
of the small circle could be generated,
but any deformation of this intensity
would probably result in full flattening
onto the axial surface. Stereograms
generated from conical folds therefore
generally look like those of Figs 120 &
121.
Figure 120

EPSC 203 (2012) p. 110

In these figures, B is the new fold
axis. The dotted line marks the small
circle. The shaded regions are those
occupied by poles to bedding or
whatever surface is being folded. In Fig.
120 the distribution of poles is so small
that you could fit almost anything to it.
In Fig. 121 you could fit a great circle to
the data, without realizing there was a
conical fold.
For cylindrical folds, the tightness
is measured using an interlimb angle.
The interlimb angle may also be viewed
as a measure of the extent of rotation of
the poles about the great circle. Thus,
the tightest folds, with interlimb angles of
00, are produced by rotation of the poles
Figure 121
through 1800 about the axis, on the
great circle. An interlimb angle of 900 reflects a rotation of 900; an interlimb angle of 450
reflects a rotation of 1350; and so on.
For a conical fold, the same feature - extent of rotation about the axis - is probably the
best measure of tightness and, of course, the conical angle should be given.

EPSC 203 (2012) p. 111

RELATIVE AGES OF FABRIC ELEMENTS

It is possible to recognize the relative

ages of fabric elements by their stereogram
patterns in many cases. It is rare, however,
for the stereogram pattern alone to provide
unambiguous interpretations. It is crucial to
make observations in the field concerning
the relative ages of fabric elements. The
most important and useful criteria, which can
generally be worked out with a little common
sense alone, are:
Figure 122
(a)
S2 offsets S1. i.e., careful observation of hand specimens (with a hand lens in many
cases) reveals that individual S1 surfaces have been offset on S2 (Fig. 122).

(b)

S2 is the attenuated limbs of folds in

In this case S2 may be
S1.
considerably
enhanced
by
metamorphic differentiation (Fig.
123).

Figure 123

EPSC 203 (2012) p. 112

(c)

S2 is defined by the preferred

orientation of grains whose spatial
This
distribution determines S1.
criterion may require microscopic
examination, and therefore the
collection of oriented specimens (Fig.
124).
Figure 124

(d)

Planar S2 cut microscopically folded

S0 whose axial planes are S1 (Fig.
125).

Figure 125

EPSC 203 (2012) p. 113

INTERFERENCE PATTERNS IN PLANAR SECTION
Interference patterns provide one of the most powerful means of unravelling the
deformation history of complexly deformed regions. They may be observed on all scales,
from several kilometres down to the microscopic. They are simply sections through refolded
folds. Three major types of pattern are possible, plus various combinations.
(a) COAXIAL PATTERN ('Type 3' pattern of Ramsay)
A Type 3 pattern is produced when
B1 is parallel or nearly parallel to B2, and the
axial surfaces for the two sets of folds are
orthogonal (Fig. 126).
The relative ages of the two fold sets
are easily recognized, since the axial
surfaces of F2 folds are now folded. Note
that although the ideal pattern is produced
by perfectly parallel B1 and B2, and
orthogonal S1 and S2 the same style of
pattern may be produced with the angle
between B1 and B2 as much as 450, and the
angle between S1 and S2 as little as 450,
provided the outcrop section is favourable.
Figure 126. Ramsay's Type 3 Interference
Pattern.
(b) DOME-AND-BASIN PATTERN ('Type 1' pattern of Ramsay)
A dome-and-basin pattern is produced when B1 is perpendicular to B2, and S1 is
perpendicular to S2, and the section is cut at a shallow angle to the fold axes (Fig. 127).
Note that again this is the ideal pattern. The same style of pattern would be produced
with the angle between B1 and B2 as little as 450, and the angle between S1 and S2 as little as
450, provided the section were cut at a shallow angle to the fold axes.

EPSC 203 (2012) p. 114

From the diagram it is clear that,
given a perfect dome-and-basin pattern, it is
impossible to tell the relative ages of the two
fold sets from the outcrop pattern alone,
since neither axial surface is deformed.
Stereogram patterns for the refolded fold
axis will not help; both fold axes will be
distributed about great circles. Examination
in the field of the relative ages of S may
provide the answer, as will carefully chosen
domainal plots of bedding - those closures
formed in D2 will generally be conical, not
cylindrical.
Imperfect dome-and-basin patterns
are easier to analyze, since one of the axial
surfaces is refolded by the other. i.e., if B2 is
not perpendicular to S1, S1 is folded
(conically) about B2, and the folding is visible
in the hinge-surface trace (Fig. 128).

Figure 127. Dome-and-basin interference

pattern

Figure 128. Asymmetrical dome-and-basin

pattern.

(c) MUSHROOM PATTERN ('Type 2' pattern of Ramsay)

A mushroom pattern is produced when the angle between B1 and B2 is high to
moderate (usually > 450) and the angle between S1 and S2 is high. In addition, and most
important, S1 must have lain at a shallow angle to the surface of the section. In the case of a

EPSC 203 (2012) p. 115

horizontal section this means that F1 folds must have been near recumbent (Fig. 129).

Figure 129. Mushroom pattern.

Note that the geometric conditions for a "mushroom" pattern are similar to those for a
dome-and-basin pattern, except for the angle between S1 and the outcrop surface. It follows
that the type (style) of interference pattern is a function of the attitude of the exposure as
much as the geometric relationships between the folds.
The mushroom pattern may be analyzed quite simply: the nature of the pattern is
such that F1 closures are easily distinguished from F2. In addition, it is possible to tell in which
direction the F1 folds closed before F2. To do this we need to know whether the F2 folds are
antiformal or synformal, which can be determined, if their general plunge direction is known,
simply from their senses of closure on the section. Once it is known which F2 traces are
antiforms the direction of closure of F1 folds may be determined (Fig. 130). The four general
types of mushroom pattern (they may be checked with paper models quite quickly) are the
only ones possible.

EPSC 203 (2012) p. 116

Figure 130. Schematic of the four possible combinations giving mushrooms.

If we also know which F1 folds are anticlines, i.e., if we know the order of the
stratigraphic succession, we can say in which direction the F1 folds were vergent before F2.
e.g., a south-closing F1 anticline requires south vergent F1 folds, etc. (Fig. 131).
It is important to
emphasize
that
the
geometric
conditions
specified for each type of
interference pattern are not
tightly constrained; they may
be relaxed significantly while
still maintaining the same
style of pattern. In addition, it
is important to realize that
Figure 131. South-vergent folds; the anticlines close
these three types are only
southwards.
end-members.
Various
intermediate
types,
and
combinations of types, are possible given suitable geometry.

EPSC 203 (2012) p. 117

SHEAR-ZONE DEFORMATION
Deformation in shear zones is distinguished from regional-scale deformation by the
fact that deformation is not homogeneous on a large scale; i.e., the shear zone itself is a
high-strain zone, and it passes outwards, either discretely or progressively, into a zone of no
strain. Whether the passage is discrete or progressive depends largely on the conditions
under which the shearing occurred - ductile shear-zones, formed at high temperatures, tend
to pass progressively out into undeformed rocks, whereas brittle shear zones may have
sharper boundaries.
Some insight into the structures that may be expected in shear zones may be gained
by considering the simplest possible type of shearing, homogeneous simple shear. In
homogeneous simple shear the body is subjected to a shearing couple, giving rise to
progressive rotation and flattening (Fig. 132).
At

any given
stage
during
the
shearing,
the
instantaneous plane of
flattening is at 450 to
the boundary of the
shear zone, and is
perpendicular to the
plane of extension.
This means that any
response of the rock to
that
instantaneous
strain, such as the
Figure 132
opening of fractures
perpendicular to the maximum principal stretch or the development of cleavage perpendicular
to the minimum stretch, will be governed by this orientation. During the progressive strain that
accompanies motion on the shear zone, however, any early-formed features are rotated.
This may result in a sigmoidal form to the features, they having experienced greater rotation
in the centre of the shear-zone, where strain is generally greatest, and lesser rotation near the
edges. Under some circumstances, response to late-stage instantaneous strain may produce
extensional fractures that transect earlier-formed fractures that have been rotated during the
progressive shear (Fig. 133).

EPSC 203 (2012) p. 118

Deformation in
shear zones involves
the progressive displacement of tabular
domains within the
zone, in a direction
parallel to the shear.
This
motion
corresponds precisely
to the mechanism of
shear folding.
It is,
therefore, possible to
Figure 133
produce ideal similar
folds, with all the
characteristics of classic shear folds, in a shear-zone environment. A further type of fold that
has commonly been recognized in shear zones is a sheath fold. Sheath folds are folds with a
broadly conical form; i.e., they have the form of a cone (they resemble the sheaths of
daggers) rather than being draped around a cone, as is the case for conical folds. Generally
the long axes of the sheaths point in the direction of shear. The mode of origin of sheath folds
is uncertain. It is possible that in some cases they form by the superposition of
inhomogeneous strain on pre-existing folds. It is thought, however, that some sheath folds
nucleate and develop entirely as a consequence of the shearing, perhaps by nucleating on
the terminations of relatively unsheared lozenge-shaped blocks bounded by zones of high
shearing strain.
As illustrated above, flattening during progressive shear may produce a cleavage at
roughly 450 to the plane of the shearing. The plane of the shearing may itself be represented
by a surface of discontinuity. This gives rise to two intersecting planar fabrics, the S (for
schistosit) and the C (for cisaillement - French for shearing). S-C fabrics may be well
developed in shear zones, and provide a reliable indication of the direction of motion on the
shear-zone (Fig. 134).

Figure 134. C-S Fabric

EPSC 203 (2012) p. 119

DEFORMATION MECHANISMS AND MICROFABRIC
DEFINITIONS
The following are definitions of some terms commonly used in the description of
fabrics:
FABRIC: The internal geometric configuration of the elementary parts of a rock. The
"elementary parts" of the rock are the individual crystals. Thus, their spatial arrangement or
mutual relationships constitute the fabric (or "microfabric"). The term "fabric" may also be
used to describe mesoscopic features such as cleavage.
FABRIC DOMAIN: any domain in which the fabric is homogeneous.
FABRIC ELEMENT: any individual geometric feature that constitutes a part of the fabric.
e.g., the orientation of mica platelets (for a microfabric), an intersection lineation (for a
mesoscopic fabric).
SUBFABRIC: a fabric due to one fabric element only. The total fabric is the sum of all
subfabrics.
CRYSTALLOGRAPHIC FABRIC ELEMENT: a fabric element defined by crystallographic
rather than dimensional characteristics of mineral species; e.g., the c-axes of quartz grains.
Such a fabric element might exhibit a preferred orientation (anisotropic fabric) in the absence
of any preferred orientation to the long axes of mineral grains. This would be referred to as a
"crystallographic preferred orientation" as distinct from a "dimensional preferred orientation".

FABRIC SYMMETRY

Fabrics may be described in terms of their "symmetry", in a system similar to that used
in crystallography. Only five types of symmetry are important in fabric analysis:
(a) Spherical: the fabric may be reflected through any mineral plane in any orientation. Such a
fabric is isotropic, and is very rare, occurring only in some igneous intrusions and hornfelses.
(b) Axial: the fabric may be reflected through any of an infinite number of mirror planes, all of
which have a common axis. Examples of such a fabric are : a single lineation (axis is parallel
to lineation) and a single foliation (axis parallel to pole to foliation).
(c) Orthorhombic: the fabric may be reflected through only three orthogonal mirror planes.
Examples are: a single lineation lying in a foliation, two orthogonal foliations.
(d) Monoclinic: the fabric may be reflected through only one mirror plane. e.g., a rock with a

EPSC 203 (2012) p. 120

foliation, and a lineation not lying in it; two non-orthogonal foliations.
(e) Triclinic: there are no mirror planes.
It is obvious from the above that most natural fabrics in tectonites will be triclinic. The
subfabrics, however, may have much higher orders of symmetry.
ORIGIN OF FABRIC
Fabric may be:
(a) Of primary origin; e.g., flow lineation in a sedimentary or igneous rock.
(b) Mimetic = inherited from some primary or pre-existing fabric but preserved or enhanced
during renewed mineral growth or re- equilibration to a new mineral assemblage. e.g., the
growth of micaceous minerals during metamorphism of a siltstone, by nucleation of new
micas on old clay grains, so that the preferred orientation of micas mimics the original
preferred orientation of the clay minerals.
(c) Developed during deformation. The fabric is the result of recrystallization and/or
neocrystallization, and crystal rotation during deformation. A rock whose fabric is produced in
this way is referred to as a "tectonite", and the fabric is a "tectonite fabric". Most deformed
rocks have tectonite fabrics.

EPSC 203 (2012) p. 121

The figures for the rest are not scanned; perhaps not available.

MECHANISMS PRODUCING TECTONITE FABRICS

The mechanisms producing tectonite fabric are the mechanisms by which the rock
deforms on the grain-size scale. There are several such mechanisms:
(a) Body rotation, i.e. the rotation of individual grains so that their long axes have a preferred
orientation parallel to the long axis of the strain ellipse. This is clearly important in the
deformation of unconsolidated rocks, and may lead to the development of cleavages in some
slates, for example. It may also be important in consolidated rocks, but in this case the matrix
must be recrystallizing at the same time, and, if the grains are in contact with each other,
there must be grain-boundary diffusion, effecting dissolution of grain edges that impinge on
one another. This strain mechanism, in which there is no internal change in the grains, but
changes in shape due to grain-boundary dissolution, is sometimes known as "Coble Creep".
Body rotation may produce strong dimensional preferred orientations, which may or may not
be crystallographic.
(b) Intragranular flow., i.e., adjustments within grains in response to an external stress field.
They take place by the movement of dislocations through the lattice, or by the diffusion of
vacancies through the lattice. At low temperatures (relative to the melting point of the mineral)
dislocations can move only along certain crystallographic planes in a given mineral species.
This is known as translocation gliding . As a result they "stack up" against each other, and it
becomes harder to deform the grain with increasing strain ("work-hardening"). This effect is
reduced at higher temperatures, because dislocations are able to jump up through the lattice
and annihilate each other. The jumping of dislocations is known as dislocation climb. It allows
almost unlimited deformation of a mineral grain, and is a very important mechanism producing
preferred orientations, if it is accompanied by intergranular flow (because of impingement
problems) (Fig. A, next page).
Movement on the glide planes (sinistral in this case) is accompanied by grain rotation
(dextral in this case) so that the net effect is to rotate the pole to the glide plane towards 1.
Diffusion of vacancies through a lattice, which is a solid diffusion mechanism, is
important only at high temperatures. It is known as Nabarro-Herring creep.
Twin gliding is a special case of intragranular flow.
When a solid rock is deformed, all mechanisms are active: Coble creep, translation
gliding, dislocation climb and Nabarro- Herring creep. Generally, however, one mechanism is
much more effective than the others, under given physical conditions, so that it dominates the
deformation. The control of physical conditions on the type of mechanism is commonly
represented by "deformation maps".

EPSC 203 (2012) p. 122

DEFORMATION MAPS
The commonest type of deformation map is a plot of / versus T/Tm, where =
deviatoric stress, = shear modulus, T = temperature, Tm = melting temperature for the
mineral. T/Tm is known as the "homologous" temperature. The general form of the map, for a
given mineral species at a certain grain-size, is shown on Diag. B of the next page, in which:
dashed lines are contours of
DC = dislocation climb
NH = Nabarro-Herring creep
CC = Coble creep
LT = Low temperature creep (too low for climb)
Effect of grain size :
Larger grains deform more rapidly by dislocation climb. An increase in grain-size
therefore expands the DC field, and shrinks the CC and NH fields.
Strain of mineral grains produces a build-up of dislocations and general lattice
imperfections. "Recovery" is the removal of these imperfections. It may take place only at
reasonably high temperatures (i.e., above the LT field) where dislocations can climb. It
results in the development of "clean" grains, and commonly leads to the break-up of large,
strained grains into a mosaic of smaller grains with slightly different orientations.
RECRYSTALLIZATION & NEOCRYSTALLIZATION
(a) Recrystallization = adjustment of crystal size, shape etc., maintaining existing mineral
species, i.e. without development of new mineral species, but developing new mineral grains.
Grains that have unfavourable orientations with respect to the stress field, and are therefore
undergoing strain, are selectively removed. New, nucleating grains develop only in
favourable orientations.
(b) Neocrystallization = recrystallization in which the new grains formed belong to a new
mineral species, e.g. in conditions of prograde metamorphism. Again, the new nucleating
grains will grow only if they are favourably oriented with respect to the stress field.
These processes lead to very strong crystallographic preferred orientations.

SYMMETRY PRINCIPLE
In practice it is difficult to determine which mechanism or mechanisms have been
dominant in the deformation of a rock. Dislocation mechanisms may leave traces in the form
of deformation lamellae, but they may be eliminated by the recovery process. The major part
of deformation may take place during a metamorphic event. The new mineral species formed

EPSC 203 (2012) p. 123

in this event may then record only the post-metamorphic deformation, which may be relatively
minor. However, rocks such as calcite-rich limestone (marble), quartzite and peridotite, which
do not change their mineral assemblages during deformation under most reasonable physical
conditions, can be studied to determine which mechanisms were operative and, in the case of
dislocation climb, which slip systems were operative. This is an important field of current
research.
Even without this understanding, the character of the new fabric may indicate
something about the character of the strain and stress, using the following principles:
(a) The symmetry of a fabric can be of no higher rank than the symmetry of the least
symmetric of its subfabrics.
(b) If the subfabrics of a rock are not homotactic (i.e., do not have common axes) the
symmetry of the fabric will be of lower rank than the symmetries of the subfabrics.
(c) The symmetry of the fabric is a result of the pre-deformation symmetry and the symmetry
of the movement pattern during deformation. The movement pattern is the sum of all
componental movements in the body, i.e. the sum of the paths by which each particle moves
during the change in shape.
e.g. , the change in shape from Diag. A (next page) to Diag. B could be effected by a simple,
progressive flattening, in which case the componental movements would be of the form
shown in Diag. C (next page).
This is a "pure shear", and has orthorhombic symmetry.
Alternatively, the change in shape could be effected by two
superimposed simple shears (Fig. D, next page).
In this case componental movements would have a very different form (Fig. E, next page).
This movement pattern has only monoclinic symmetry.
These are just two examples. For any given total strain there is an infinite number of
possible movement patterns, although there are not many probable ones.
(d) The fabric produced by a movement pattern will have a symmetry rank no lower than that
of the movement pattern itself. In order for this not to be true there would have to be some
mechanism for the introduction of asymmetry. This could only be some aspect of the
movement pattern. Hence we arrive at a contradiction.