Understanding Rock Stress

John A Hudson

Lecture 2 1
 Rock masses already contain rock stresses.

Do we understand rock stress? Why do we need to understand it?

And can it be reliably determined at a particular site?

Here are the fundamentals…hopefully presented

In a relatively simple way.

It’s essential to understand stress to work

in rock mechanics and rock engineering

2
 The basic motivations for understanding in situ rock stress are two-fold.

1. To have a basic understanding and knowledge of the stress state for

interpreting structural geology features and for engineering analyses.

How did the rock become deformed and/or fractured in that way, and
for engineering what stress effects are we defending ourselves and
our structures against?

2. To have a specific and 'formal' knowledge of the boundary conditions

for stress analyses conducted in the design phase of rock engineering
projects.

There are many cases in rock engineering where the stresses are not
applied as such; rather, the stress state is altered by the engineering
activities, e.g. in the case of excavating a rock slope or tunnel.

The rock is currently pre-loaded, i.e. pre-stressed

3
 The generic
rock mechanics/
rock engineering
problem

F3 Boundary
F2
conditions
F1
Fn
Intact rock Excavation

Water flow

Fractures

So, let’s think about rock

stress in the context of our
generic diagram.
4
 The rock mechanics and rock engineering problem

F3 Boundary
Characterising the F2
conditions
forces on and F1
throughout the rock
mass through the Fn
concept of stress Intact rock Excavation

Water flow

Fractures

What is the stress at a point

within a rock mass?

5
Point 1: The need to understand rock stress

Stresses in rock masses have caused the

geological structures that we see today and
knowledge of these is required for the modelling
and design of current engineering projects.

During rock engineering, the pre-existing natural

rock stress state can be significantly altered.

Stress is not a familiar concept.

6
Point 2: Stress is not a scalar or vector
quantity, it is a tensor quantity

A scalar quantity has magnitude only, e.g.

temperature, energy (one component required for
its specification, e.g. 6C).

A vector quantity has magnitude and direction,

e.g. velocity, force (three components required for
its specification, e.g. 6 N acting along a line with
trend 234 and plunge 6).

A tensor quantity has magnitude, direction and the

plane that it is acting on, e.g. stress and strain (six
components required for its specification)
7
 A solid can sustain a shear force, whereas a liquid or
gas cannot.

A liquid or gas contains a pressure, i.e. a force per unit

area, which acts equally in all directions and hence is a
scalar quantity.

Rock can sustain a shear force and hence contains

stress, which is a tensor quantity.

8
Point 3: There are normal forces and there are shear
forces, and there are normal stress components
and shear stress components.

The force can be resolved into a normal component, Fn, acting

perpendicular to the plane, and a shear component, Fs, acting
parallel to the plane.

A normal stress, , can be resolved into a normal stress

component, n, and a shear stress component, . However, while
the resolved force perpendicular to the plane Fn = Fcos, the
resolved stress perpendicular to the plane, n = cos2 because
the stress quantity is related to force/area: while cos is sufficient
for the force resolution, cos2 is required for the stress component
resolution—because both the force and the area have to be
resolved in the case of stress. 9
Point 4: Stress is a point property

Considering a small cube within a rock, the normal and shear stress
components are as indicated with the notation explained at the right-
hand side of the Figure: with reference to the axes directions shown
on the left, the first subscript for a stress component indicates the
plane on which the stress component is acting; the second subscript
indicates the direction in which the stress component is acting.

These stress components are considered as limits when the volume

of the cube is reduced to zero so that stress is regarded as a point
property. 10
Point 5: The nine stress components can be
listed out in matrix form

The nine stress components can be listed out in a table form.

This is known as the stress tensor. It is a more complex
quantity than a scalar or vector because the resolution of all
the components when the cube is rotated is more
complicated than simply resolving forces.

11
Point 6: The stress matrix is symmetrical.

Some of the off-diagonal terms in the matrix are

equal. This is a result of ensuring that the cube does
not rotate. In the x-y plane shown above, the cube
will not rotate if xyl = yxl, in other words if xy =
yx, and similarly for the x-z and y-z planes, xz = zx
and yz = zy. 12
Point 7: The state of stress at a point has six independent
components.

As a result of the equality of shear stress components in each plane,

only six of the nine components in the stress matrix are independent.

The consequence is that the state of stress at a point has six

independent components with the corollary that the stress at a point in a
rock mass can only be specified by designating the six components.

These can be the six components in the matrix given with reference to a
set of axes or, more usually, they are given with reference to the
magnitude and orientations of the principal stresses (see the next Point).

If less than six components are given, the statement has no meaning,
e.g. the stress is 6 MPa has no meaning because six components are
required to specify the stress state.

13
Point 8: There is an orientation in space for which all the shear
stresses vanish and there are only normal components of
stress—the principal stresses.

It is perhaps surprising that, given the existence of the six shear stress
components, it is always possible to find an orientation of the cube for which
all the shear stresses disappear leaving only normal stresses acting.

These normal stresses are known as the principal stresses, the major,
intermediate and minor principal stresses: 1, 2 and 3 respectively.

The converse is not true: we cannot find an orientation of the cube such that
all the normal stresses disappear leaving only principal shear stresses. 14
Stress analysis can become complicated, and is beyond the scope of
these lectures…

15
Point 9: All free surfaces are principal stress planes.

Whilst there can be shear stresses within a solid, there can be no

shear stresses on a free surface—because there can be no
reaction to sustain the shear stress.

It follows immediately that a free surface is then a principal stress

plane, i.e. the small cube must be orientated with one face being
the free surface.

Moreover, there is also no normal stress acting on a free surface

and so the associated principal stress acting perpendicular to the
surface is zero.

1
 1 0 0 2
0  0
,
 2

 0 0 0 3 = 0

16
Principal stresses are parallel and perpendicular
17
 1
 1 0 0 2
0  0
,
 2

 0 0 0 3 =

Principal stresses are parallel and perpendicula

to open fracture surfaces and excavation surfac

Influence of an open fracture on the local stress state

18
Plotting the stress state

19
Diagrams thanks to Prof John Harrison

20
Point 10: Measuring rock stress is not an easy task

This is due to a combination of adverse factors, including the facts that

• stress is a point property,

• the stress state must be altered in order for the measurements to be made,

• at least six measurements are required,

• local measurements may not indicate the overall stress state, and

• the stress measurements are often made remotely in boreholes.

21
Natural stress: the in situ stress which exists prior to engineering.

Induced stress: the natural stress state as perturbed by engineering.

Gravitational stress: the stress state caused by the weight of the rock
above.

Tectonic stress: the stress state caused by tectonic plate movement.

Residual stress: the stress state caused by previous tectonic activity.

Thermal stress: the stress state caused by temperature change.

Palaeostress: a previous natural stress that is no longer acting.

Near-field stress: the stress state in the region of an engineering

perturbation.

Far-field stress: the stress state beyond the near-field.

Local stress: the stress state in a region of interest. 22

Different scales

• Tectonic scale and

regional stresses

• Site scale

• Excavation scale

• Borehole/measurement scale

• Microscopic scale
23
23
Stress Perturbation Factors
The
• Rock inhomogeneity

• Rock anisotropy

• Rock discontinuities

• Influence of a free surface

1 0 0 Principal stresses parallel to excavation

  surface 1
0 2 0 .
  Principal stress perpendicular to 2
 0 ,
0 0 excavation surface
24
24
3 = 0
Influence of a fracture zone on the local rock stress: AECL URL, Manitoba

Diagram thanks to Prof Derek Martin

25
From “Geological Engineering”
by Gonzalez de Vallejo & Ferrer

26
Case Example of In Situ Stress

(next four slides from Max Lee - AMC

Consultants Pty Ltd, Australia)

27
28
 Principal Stress Magnitudes versus Depth
Magnitude (MPa) Magnitude (MPa)
0 20 40 60 80 100 120 0 20 40 60 80 100 120
0 0
1 1
2 2
200 3 200 3

400 400
Depth Below Surface (m)

Depth Below Surface (m)

600 600
Magnitude (MPa)
0 20 40 60 80 100 120
0
1
2
200 3

400

Depth Below Surface (m)

600

800

1000

1200

800 800

1000 1000

1200 1200

Yilgarn Craton Eastern Australia

Note the spread of results in each case.

Wuhan – Feb 2008 – Unsolved Problems in Rock Mechanics and Rock Engineering 29
 I = 1 + 2 + 3
Principal Stress Magnitudes vs 1st Stress Invariant
Principal Stress Magnitude (MPa) Principal Stress Magnitude (MPa)
0 20 40 60 80 100 120 0 20 40 60 80 100 120
0
1 1
2 2

50 3 50 3

Sum of Principal Stress Magnitudes (MPa)

Sum of Principal Stress Magnitudes (MPa)

100 100

Magnitude (MPa)
0 20 40 60 80 100 120
0
1
2
200 3

400

Depth Below Surface (m)

600

800

1000

1200

150 150

200 200

1 /1 /2 2 1.5

1.5 1 1/ /2 2 1.5
1.5
250 250
 2 /  3  1.5
2 / 3  1.5 2 2/ /3 3 1.5
1.5

 1 /  3  2.3  1 /  3  2.3
1 / 3  2.3 1 / 3  2.3
300 300

Yilgarn Craton Eastern Australia

Note how
Wuhan – the– Unsolved
Feb 2008 stress Problems
trendsinhave been clarified
Rock Mechanics by plotting I against depth.
and Rock Engineering 30
 Principal Stress Magnitudes vs 1st Stress Invariant:
Combined Data
Principal Stress Magnitude (MPa)
0 20 40 60 80 100 120
0

Yilgarn + Eastern Australian data

50
Sum of Principal Stre s s Ma gnitude s (MPa)

• Same relationship!!

100
• Two hypotheses

 Structures control the principal stress

ratios that rock masses can sustain,
150
irrespective of how strong or weak the intact
rock might be
200
 It’s likely that all rock masses try to sustain
the maximum “geologic” stress that they
250
1 / 2  1.5 can sustain
2 / 3  1.5
1 / 3  2.3
300

…and that the same relationship emerges for West and East Australia.
Wuhan – Feb 2008 – Unsolved Problems in Rock Mechanics and Rock Engineering 31
 Your lecturer in Japan with three
world experts on rock stress
and its measurement! The

2003

FairhurstBezalel Haimson Jean-Claude Rogiers John Hudson

Colloquium, 15-16 October 2004, Minneapolis Francois Cornet 32
32
Issue of this journal
devoted to rock
stress estimation
which contains the
four International
Society for Rock
Mechanics
Suggested Methods
and 17 papers on
experiences of
stress measurement

33
 F3 Boundary
F2
conditions
F1
Fn
Intact rock Excavation

Water flow

Fractures

ISRM Suggested Methods for Rock Stress Estimation

Part 1: Strategy for rock stress estimation
Part 2: Overcoring methods
Part 3:Hydraulic fracturing and/or hydraulic testing of
pre-existing fractures (HTPF) methods
Part 4: Quality control of rock stress estimation

The Suggested Methods have been published in a Rock Stress Estimation

Special Double Issue of the International Journal of Rock Mechanics and
Mining Sciences (Volume 40, Issue 7-8, 2003), together with a suite of
supporting papers on various aspects of establishing the rock stress state.

34
End of Lecture 2

35