You are on page 1of 484

ods of

Geoogica
• •
9 neer.ng
in Discontin ous Rocks

RICHARD E. GOOD N
Professor of Geological Engineering
University of California, Berkeley

WEST PUBLISHING COMPANY


ST. PAUL • NEW YORK • BOSTON • LOS AN GELES • SA N FRANCISCO
COPYRIGHT © 1976 By WEST PUBLISHING CO.
All rights reserved
Printed in the United States of America
Library of Congress Cataloging in Publication Data

Goodman, Richard E.
Methods of geological engineering in discontinuous rocks.
Includes index.
1. Rock mechanics. I. Title.
TA706.G66 624'.1513 75-42152

ISBN 0-8299-0066-7
v

preface

This work contains material from my courses at Berk eley in

Engineering Geology and Applied Rock Mechanics. It could serve as a

supplementary source for general courses in these fields, as well as

a text in a more specialized course in geological engineering analysis.

I have not tried to separate the disciplines of geology and civil

engineering as I think it unnatural to do so. The geologist can

decide on the relevancy of the features he maps and describes only

if he understands how they relate to the analytical process. The

engineer can not handle the geological data correctly without the

respect for its !! fuzziness II that comes from a personal acquain tance

with geology.

Interest in geological engineering methods has occupied me since

I visited the sites of the Malpasset and Vajont failures in the early

60 1 s. I wanted to know if a person with reasonable education and

experience in engineering geology could have foreseen the difficulties.

It seemed to me that objective, rational procedures for evaluating

such sites were inadequate and that we were, perhaps, relying too
vi Preface

heavily on intuition. Wisdom derived from real experiences will

always be an important and necessary ingredient for predicting rock

behavior; but we must also search for objective tools. Since 1960,

a number of such tools have appeared and it is timely and useful to

set them forth for students, and for practicing technicians who can

apply them in their work.

Most of this Monograph was written at Imperial College, London,

during the term of a Guggenheim Fellowship, while I was on a sabbatical

leave from Berkeley. I wish to thank the John Simon Guggenheim Founda-

tion and the Regents of the University of California for this opportu-

nity to reflect on the whole of the field of geological engineering

methodology. In residency at the Royal School of Mines during 1973, I

was privileged to frequent discussions with Professor Evert Hoek, Dr.

John Bray, John Boyd, and other faculty members. And I profited from

the work of a number of their students, past and contemporary, includ-

ing Peter Cundall, Christopher St. John, Nick Barton, Tidu Maini, John

Franklin, John Sharp, Laurie Richards, Dermot Ross-Brown, Peter Riley,

Ross Hammett, Peter Kelsall, Don Moy, Tim Harper, and Graeme Major. I

am particularly indebted to Dr. St. John who prepared the special

finite element program listed in Appendix 1. We were anxious to have

a small finite element program written expressly for a student trying

to bridge between theory and application.

In the text, I have indicated the sources for ideas by references

to the works of numerous authors. I have profited from personal com-

munications over the years with a number of these authors - - Walter

Wittke, Klaus John, Pierre Lande, and "Skip" Hendron with respect to

limit equilibrium analyses - - Dr. Leopold Muller and Karel Drozd with
Preface

respect to p h y sical models E. J. Polak, Tor Brekke, and Dan Moye

wi t h respect to geological and geophysical exploration - - and Robert

Tay l o r , Edward Wil son , Hugh Trollope, and Ann Bornstein with respect

to comp u ter me t h o ds. I have also benefitted from the wo rk of past and

p re s en t students at Berkeley inc l uding Yuzo Ohnishi, P. N. Sundaram,

Ashra f Mah tab, Rudolfo de la Cruz, John Cadman, Jacques Dubois, Alain

de Rouvray , and Francoi s Heuze, and of post Do ctoral scholars Kemal

Ergu vanl i, J ean Lu c Dessenne, and Karel Drozd. The critical comments

and sugge s ti o ns of P r of . Arvid Joh ns on of Stanford University, and

Pro f . Hendron and hi s colleagues at the University of Illinois were

quite h elpful .

A n umbe r of p erson s kindly loaned materials or gave me permission

to ref er to t he ir work. These include Nick Barton, Z. T. Bieniawski,

Tor Brek k e , W. Chin n , Lloyd Cluff, Phillip Cole, James Coul son, Walter

Day , Wi lliam De a r ma n, G. Everling, Irving Fatt, Alena Gralewska-Vickery,

Ri chard Ha y , Francois Heuze, Larry James, Dennis Lachel, Branko Ladanyi,

Thomas Lan g , P ierre Lon d e, Ken Ma tthews, J. My ung, Carlos Ospina, Marc

Pan e t, M. Pop o vic , Howard Pratt, Hernando Quijano, Doro t h y Radbruch,

Ni ck Ren ge r s , Manuel Ro c ha, Fritz Ru mmel, F. Sabarly, Ed. Slebir,

Ger a rdo Tarna, Jos e Tejada , Ruth Ter z aghi, Cl. Tourenq, Lloyd Underwoo d ,

a n d J oel Ver di e r.

The f ollow i n g o rgan izations generously permitted me to refer to

or borrow thei r ma t e ri als: Atlas Copco ABEM ( S tockholm), Bergbau

Forschung (Essen) , Birdwe l l Division of Seismograph Service Corp.

(Tulsa ) , Californi a Department of Water Resources (Sacramento),

Christensen Diamond P ro d u c ts Co . (Salt Lake City), Coyne and Bellier

( Par is ) J Engineering Laboratory Equipment Ltd. (Hemel Hempstead, U. K.),


viii Preface

Golder and Brawner and Assoc. (Vancouver), Ingetec Ltda. (Bogota),

Integral Ltda. (Medellin), Laboratoire des Ponts et Chaussees (Paris),

Laboratorio Nacional de Engenharia Civil (Lisbon), Longyear Co.

(Minneapolis), Joy Manufacturing Co. (Montgomeryville, Pa.), Mindrill

Ltd. (Melbourne), Norwegian Geotechnical Institute (Oslo), Soil

Mechanics Equipment Co. (Glen Ellyn, Ill.), Sprague and Henwood, Inc.

(Scranton, Pa.), Tacoma City Light, TerraTek (Salt Lake City), U. S.

Army Corps of Engineers (Libby Resident Office; Missouri River

Division, Omaha District, and Explosives Excavation Research Lab.),

U. S. Bureau of Mines, (Denver), U. S. Bureau of Reclamation (Denver),

U. S. Geological Survey (Menlo Park), Woodward Clyde and Associates

(Oakland), and Zavod Za Geotehniku I Fundiranje (Sarajevo).

Finally, I wish to thank the persons who helped me with the work

of producing the manuscript: Fran Riley, Laurie Wilson, Gloria

Pelatowski, and Lillian Goodman.


ix

to the memory of

Parker D. Trask
j
I
xi

contents

PREFACE

1. II"TRODUCTI ON

2. ROCK CLASSIF IC ATI ON 14


The Nature of Rock 14
Rock Specimen Versus Rock Mass 15
Petrologic Classification of Rock Specimens 16
Rock Versus Soil and Weathered Rock 19
Weathering 22
Index Tests for the Quality of the Rock Material 30
Fissured Rocks 30
Discontinuities 40
Continuous and Discontinuous Rock Masses 47
Engineering Classifications of Rock Masses 49

3. PRII"CIPLES OF STEREOGRAP HIC PROJECTI ON AND JOINT SU RV EYS 58


Conformal Stereographic Projection 58
Basic Constructions 66
Joint Surveys and Statistics on the Sphere 83
Bias in Measurement of Joint Orientations from Drill Holes and Outcrops 86
The Directionality of a Jointed Rock Mass 88

4. EXPLORATION OF ROCK CON DITIOI"S 91


Geological Maps and their Interpretation 91
Aerial Photo Interpretation 104
Terrestrial Photographs 112
Geophysical Methods 121
Drill Holes 127
Absolute Orientation of Structural Features in Drill Core 142
xii Contents

5. MECHANICAL PROPERTIES OF DISCONTINUITIES 158


Determination of Properties 158
Deformations in Joints 170
Peak Shear Strength 183
Influence of Joint Orientation 200

6. APPLICATIONS OF STEREOGRAPHIC PROJECTION IN


MECHANICS OF DISCONTINUOUS ROCKS 269
Introduction 209
Kinematical Considerations 210
Operations with Vectors on the Stereonet 217
Application of the Stereographic Projection in Defining a Wedge
Formed by Intersecting Discontinuities 225
Analysis of Rotation 231
Analysis of Sliding of a Block on a Plane-the Friction Circle Concept 237
Estimate of the Displacements of a Block Under a Dynamic Impulse 244
Slip of Tetrahedral Wedges 247
Sliding of Tetrahedral Wedges With Only One Free Surface 255
Slides Composed of Two Blocks 261
The State of Stress in Rocks 270
Conclusion 275

7. PHYSICAL MODELS 277


IKinematic Models 277
Physically Scaled Models 285

8. THE FINITE ELEMENT METHOD 300


Introduction 300
The Method 301
Formulation of Element Stiffness Matrices and External Loads 307
The Constant Strain Triangle 308
Initial Stresses in the Rock 317
Constant Strain Joint Element 320
Assembly of the Structural Equations 330
Iterative Solution to Simulate Real Properties of Joints 333
Sources of External Load 349
Example Problems 354
Incremental Loading 367
Rigid Block Analysis 368

APPENDIX ONE-AN ILLUSTRATIVE FINITE ELEMENT PROGRAM 369


Purpose and Scope of the Computer Program 369
Program Structure 371
Input Instructions 375
Examples of Input and Output 378
Listing of Finite Element Program-"JETTY" 394

APPENDIX TWO-CONVERSION FACTORS 417

REFERENCES 419

INDEX-Subject 451, Author 467


Methods of Geological Engineering
in Discontinuous Rocks
1
i troduction

Thi s b o o k d iscuss e s methods a n d pro c e du r es ava i labl e to assess


the i n fl uence o f discontinuities on the behavior of r o cks in engineer-
ing app l ica ti on s. Most rock masses in the region of influence of
works s u ch as qua r r i e s , road cuts, foundatio n s, dam abutments, tunnels,
and un der g r ound chamb ers contain planar surfaces of potential or real
weakn e ss. The se we akn ess planes come in all lengths and spacings and
have vary i ng deg r ees of i n fluence on the overall mass properties. We
rare l y can a f ford to c l os e ou r ey e s to thei r presence in attempting
to cal cu late ro c k per fo rmance.
Us e o f t h e p h rase "di s c o nti nuous rocks" in the title implies
th a t th er e are other ro cks wh ich are truly continuous. This is not
s t r i c t ly correct for even the mightiest wall of granite has exfolia-
t i on s u rfaces and other widely spaced joints and faults in various
orien t at i ons . There are ma n y rock masses, however, in which the
discon ti nui tie s t ho ugh present a r e not the we akest link in the list of
compon ent s whi ch col l e ctively give the rock its strength and other
physical at tribu t es. In friable sandstones of Tertiary age, for
examp l e , the s a n d g rai n s may be so poorly bonded that failure through
the rock mate r ial itself is more likely than failure by sliding on
bedding pl an e s or joints . This may also be the case in shales which
t hough dis r up ted and loosened by anastamosing cracks due to slaking
near exposed s u r faces, will tend to fracture through the body of
materia l rathe r t han on structural ly controlled surfaces. The y ounger,
2 Introduction

Figure 1-1. A discontinuous rock


mass. Columnar joints and flow band-
ing in a basaltic flow-Iceland;
(courtesy of Dr. Tor Brekke).

weaker rocks as a rule tend more closely to fit a "continuous model",


while hard rocks invariably are controlled in their failure modes by
their pattern of discontinuities. How else could a rock like quartzite
fail? It has a compressive strength some ten times that of mass con-
crete.
Soils are not continuous materials; they have grains and pores.
But they have been successfully analyzed using a continuum model --
heterogeneous, if necessary, but continuous. The discontinuous rocks
with which we are concerned here might at first thought be likened to
soils and treated using soil mechanics theory and techniques. In
fact, some early attempts were made in this direction. But there are
fundamental differences. The discontinuous rocks have essentially no
pore space, except that of the rock material itself (pores in the
rock material are analogous to pores in the grains of the soil). Thus
the discontinuous rock is locked together into a perfectly fitted
pattern. To create failure, pore space must be created and this
implies dilatancy, or bulking in the construction man's parlance.
Not only normal and shear forces act inside such rock masses, but
moments as well. Soil grains may be free to turn in place; rock
blocks are not.
It might seem a hopeless quest to rationalize the design
process when dealing with such a material as discontinuous rock.
Sometimes it is hopeless, and only previous experience, or trial and
Introduction 3

error, can be used. Other times, fortunately more f req u e nt l y as we


gain experience, the network of discontinuities can be accurately
described and mapped and its influence on the mass behav ior can be
adequately evaluate d . The elaboration of these methods i s the subje ct
of this book. First we must meas u r e the orientatio n s of t h e v a rio us
sets of planes which penetrate the rock in quest ion. This can be do n e
by geological observations on outcrops, by i n s p ecti on of n a tural and
artificial cuttings, by study of aerial phot ogr aphs , by me as u r ements
on drill cores and the walls of exploratory borings, and by ge op hysi-
cal traverses using a number of avail ab le techni ques. Chapt e r four
reviews some of the relevant methods of exploration aimed at p ro vi d ing
a description of the orientation and spa ci ng of the discontin ui ty
network in a rock body. Chapter five then introduces t he mechani cal
properties of surfaces of d i scontinuit y and consi de rs thei r meas ure-
ment and numerical values. When we are dealing with s in gle, very
important weakness surfaces, whose orientation and position with
respect to a project are known with precision, it is possi ble to make
explicit analyses of the resulting stresses and deformations ; this
can be done by kinematics and statics, using stereog r ap h i c pr o j ecti on
to handle the three dimensional aspect of the problem, as di s cus sed
in chapter 6. In chapter 7, physical model methods are introduced.
An emphasis is placed on kinematical models which examin e t he v a rious
possible modes of failure of a discontinuous rock mass in an engineer-
ing context. Analyses can also be performed by numer ic a l me th ods;
the finite element method is introduced in chapter 8 a nd a di g i tal
computer program, designed so that it can be read along wi t h the
theoretical discussion of chapter 8, is presented in the Appendi x.
Because stereographic projection p r o cedures are used fre quently
throughout the book as a means of solving spatial problems, su ch as
orienting planes in drill core, measuring angles on terrestrial photo-
graphs, resolving stresses on planes of given orientation, a nd operat-
ing with vectors , a chapter has been addressed specifical ly to tech-
niques of stereographic projection (chapter 3). In deali ng with
vector quantities, we must use the whole sp h e r e so the s ub je ct i s
treated somewhat differently than in works on struct ura l geology .
Chapter 2, on classification of rock, has been written to r elat e the
Good man-Geolog ical Eng. - 2
4 Introduction

d i scontinuous rocks to othe r categories of rock J i .e. to se t thi s wo rk


i n its proper context .
Geological ngin e ering is concerned with a bro a d spe ct rum of
natural proces ses. At one end of the spectrum are th os e g eologic
hazards , such as large landslides, active fau l t s, an d cave rno us
terrain , which dwarf an intended project· in terms of si ze, po tential
energy , or the cos t of neutralizing the ha z ard; wit h such h az ards,
the geological eng i neer can d o l ittle more than recog ni ze, d escribe
and be responsive to eventualities. He uses vari ous methods to
study thei r potential and t o observe the i r acti v i t y , but h e has little
effect on t he phenomena t hemselves. At t h e o ther en d of the spectrum
of geological engineering applications are min i ng a nd quarrying
ac ti vities where the geology i not onl y studi e d an d ev a l uate d , but

Figure 1-2. A concrete arch dam. Mossyrock Dam, Cowlitz River, Washington: a
doubly curved, thin arch dam 365 feet above riverbed, 606 feet above the basalt
bedrock; (courtesy of Tacoma City Light).
Introduction 5

Figure 1-3. Malpasset Dam site,


looking at the left abutment an d into
the reservoir area.

wherein the r ock is removed, crushe d , st o ckpiled and perhaps even


emplaced i n a hosti l e and caus ti c environment, for examp le , as
aggregate in cement. In between are those constructions and excava-
tions, such as dams, underground openings and open cuts, which apply
static or dynamic loads or un loads at the surface or subsurface. It
is with these that the methods discussed in this book are primarily
co ncerned.
Large dams, especially co n c rete ar ch dams as in figure 1-2
combine large loads wi t h the hydraulic and chemical effects of water
and therefore place challengi ng demands on geological engineering
investigations. Much of the recent interest in geological engineering
and rock mechanics has in fact been motivated by concern about the
safety of d ams a nd reservoirs following the catastrophes at Malpasse t
dam in France and Vajont reservoir in Italy. At Malpasset dam,
fi gu re 1 - 3, a compl ic ated set of circumstances deriving from the
behavior of the sch ist ose gneiss bedrock caused a rupture of an ar ch
dam. French investigators determined that a wedge of rock in the
abutment, bounded by intersecting weakness surfaces, moved due to the
thrust of the dam and high water pressure within the abutment (Bernaix,
196 6) *. The high wate r press ure was generated by the development of

* References will be fo un d in the Bibliography, on page 419


6 Introduction

Figure 1-4. Kukuan Dam, during construction; (courtesy of


Coyne and Bellier).

a natural flow barrier under the line of action of the dam as fissures
within the rock mass closed in response to applied load. The Vajont
failure (Muller, 1964 and 1968) occurred when a massive landslide
moved on bedding surfaces into a relatively small reservoir, causing
overtopping and flooding. The landslide was triggered by uplift
forces associated with reservoir filling.
The large influence of discontinuities on construction operations
in rock is well illustrated by the Kukuan arch dam, designed by
Coyne and Bellier for Taiwan Power Company. This dam, 86 meters high,
was constructed in a valley cut 500 meters deep into alternating
layers of slate and quartzite. Thin clay seams containing graphite
compromise the stability of unfavorably oriented layers of the site.
The right bank (figure 1-4) is a 60 - 70 degree dip slope. To
found the dam in solid rock, it was necessary to excavate through 20
to 40 meters of loosened slabs, but conventional excavation was
undesirable because of the slide potential. Grouting and "dental
work" (localized replacement of weak rock with concrete) were un-
successful. A solution was obtained in which tunnels up to 10.7
Introduction 7

meters wide were driven well into the abutments and backfilled with
concrete. Since the tunnels cut across the bedding, they were stable.
After driving a tunnel to the full depth and width, it was concreted
to within several meters of the crown. Then, after two to three
weeks, a stone protection was laid on the concrete fill and a t un nel
was excavated above. The process was repeated until eight tunnels
had been constructed, producing a stable concrete structural abutment.
Activities in advancing the construction and utilization of
tunnels and underground chambers have also created interest in methods
of geological engineering. Investigations of tunnel sites remain
fairly primitive because the sites are long, and remain inaccessible
until construction. Some attention has been focused o n assess ing the
excavatability of the rock from tests on samples, but geological and
geophysical prediction techniques, and analytical methods to forecast
formation conditions are not yet generally available.
Techniques for investigating and anaJyzing rock behavior for
underground works such as subsurface power plant chambers (figure 1 - 5) ,

(b)

(a)

Figure '·5. (a) Oroville Dam project. The dam has sh ells of gravel wh ile th e core is
derived from a vast alluvial fan; (courtesy Calif. Dept. of Water Resources).
(b) Oroville underground power station machine hall during construction. The man
standing in the lower left gives the scale; (courtesy Calif. Dept. of Water Resources).
8 Introduction

Figure '·6. Spillway excavation on left abutment of Chivor Dam, Colombia. Notice
the truck and shovel for scale. The smooth surface of discontinuity in the middle left
was exposed during construction and caused a design change. The benches are 5 meters
wide and spaced every 10 meters; (courtesy of Ingetec Ltda., Bogota).

subterranean factories, defense installations, storage chambers, and


mine shafts, on the other hand, are better developed. It is usually
feasible in such projects to make detailed investigations including
determination of rock properties, analysis, and instrumentation. An
additional aspect of investigations for underground structures not
addressed in investigations for dams, is the role of in-situ stresses.
At great depth, such as in some mines in South Africa and Canada, one
occasionally reaches the natural strength of the rock.
Surface excavations for spillways (figure 1-6), mine pits
(figure 1-7), transportation routes, power plants, and for access to
the underground, are other important areas of rock engineering. In
mines, important savings in excavation volumes can be achieved by
application of simple theory supported by field observations of geo-
logical details, back calculations of failures, piezometric measure-
ments, and analysis of the response of instruments (Hoek and Bray,
1974). Careful blasting practise and instrumentation can insure safe
operation of engineering works immediately adjacent to rock slopes,
which themselves can be regarded as engineering structures (figure 1-8).
Though the specific choices of methodology will differ among all
these types of projects, basic similarities of purpose prevail.
Introduction 9

Figure 1-7. Chamblshi Mine,


Zambia (courtesy R.S.T . Ltd.
and Prot E. Hoek) ,

First . the geo logy of the site must be defined; this entails mapping
of f i e l d exposures J study of aerial photographs, and specific explora-
t ion with excavations o r drill holes . Then . the properties of the
rocks must be assessed. Here there can be different choices of

Figure 1-8. Pre-split rock exca·


vation for Stockton Dam;
(courtesy Mr. Lloyd Underwood,
Corps of Engineers, Missouri
River Division).
10 Introduction

methods since the relevant properties to be evaluated vary greatly


according to the purpose of the project. The behavior of a complex
of underground openings reflects the initial state of stress; in
some analytically based design processes the in-situ stress will need
to be measured, or otherwise determined. In shallow rock excavations,
on the other hand, the shear strength and water pressure levels are
more critical, while for foundations the deformability of the rock is
foremost. Thirdly, through model studies, computer analysis, or
reference to appropriate similar experiences, the response of the work
at the specific site with the assigned properties is evaluated. If
unsatisfactory, the structure may be relocated or the properties may
be changed in some measure by excavation, grouting, drainage, bolting,
or other means. In this case new explorations, tests, and studies
will be inaugurated. The designer will have the most economical
solution if he is able to adapt the style of structure to the par-
ticular attributes of the site, most of which have been provided
naturally. The methods and work of geological engineering are there-
fore mainly devoted to discerning just what is already there.
The nature of rock is vastly different from other types of
engineering materials. Therefore it is natural that the methodology
employed for its characterization should be peculiar to the field of
geological engineering. Nevertheless, each of the methods employed
and discussed here has its cousins in other disciplines, and a book
such as this must cross the borders of many fields. These include
mining; petroleum; geophysics; cartography; planning; soil mechanics;
hydraulics; mechanics of materials; concrete technology; structural
engineering; statistics; aeronautics; and computer science. The
obvious consequence is that sources of literature of interest for
further reading are scattered among numerous journals, and reference
books. However, a number of basic references and journals can be
singled out as especially relevant. These are listed in Table 1-1.
Introduction 11

TABLE 1-1
Some Sources of Information

Bibliographies and Abstracts

KWIC Index of Rock Mechanics literat u re publ ished bef o re 196 9 -


2 volumes. Produced by Rock Mechan ics In formation Se rvice,
Imperial College, London.
Published by AIME, 345 East 47th St., New York, N. Y. 10017

Geomechanics Abstracts - Part II of the Inter. J o ur. Rock


Mechanics and Mining Science
Published by Pergamon Press from volume 4 ( 19 73 ) o nwa rd
(Originally called Rock Me cban i c s Ab str ac ts; pr o d uced by
Imperial College).

Geotechnical Abstracts - Monthl y with ann ua l inde x es


Deutsche Gesellschaft fur Erd- und Grun db a u ( fo r I nt e r. So c.
for Soil Mechanics and Foundation Engine e r ing).
(Published also in a ca rd format called "Geo de x Retr i e val
System " ) .

Bibliography and Index of Geology - Monthly


Geological Society of Ameri ca .

National Technical Info rma tion Service, Sp ring fie ld , Va . 22 151


(Bibliography and source for U.S. Government documents).

Geoscience Abstracts - Monthl y .


American Geological I n sti tute, Washing ton 25 D.C.
(A special supplement is devoted to a "Bi b li ograp h y of
bibliographies of the States").

Chronique d es Mines et de la Recher che Min ie re


published 10 times per ye ar by Centre d'et u des geolog iques
et Minieres

Annotated Bibliography of Economic Geology - semi - a nn u a l.


Economic Geology Publish i ng Co.

Journa ls and Serials

Rock Me chanics ( I n ter. Soc . for Rock Mechanics) Forme r ly "Rock


Mechanics and Engineering Geology" .

International Journal of Ro ck Me c hanics and Min i n g Science


(Pergamon Press).

En g ineer i n g Ge o lo g y (Elsevier).

Quarterly Journal o f Engineering Geology ( Ge ol og i c a l So c. of


London) .
12 In troduction

Bulletin of the Assc~iation of Engineering Geologi s ts.

U.S. Bureau of Mines , Reports of Investigations and other


publications.

Canadian Geotechnical Journal .

Geotechnique .

Bulletin of the Inter . Association o f Engineering Geo logy.

Proceedings of Congresses and Symposia of the International Society for Rock Mechanics*

First Congress - Lisbon 1966 - 3 vo l ume s.

Second Congress - Belgrade 1970 - 4 volumes

Third Congress - Denver 1974 - 5 volumes.

Symposium on Rock Mechanics - Madrid 1968 - 1 volume.

Symposium on Stress Measurement - Lisbon 1 970 - 1 vo l ume .

Symposium on Large Permanent Underground Opening s - Os lo 196 9 -


1 volume

Symposium on Rock Fractures - Nancy 1971.

Symposium on Percolation through Fractured Rock - Stuttga rt 1972.

Proceedings of Symposia on Rock Mechanics-U.S.A.

8th to 12th, 1966 - 1970 (AIME).

13th to 15th, 1971 - 1973 (ASeE).

Previous Symposia are l isted in preface mater i a l f or above


Symposia .

Other proceedings of interest are lis t ed in volume 2 o f KWIC


Index, (see "Bibliographies and Abstracts I I abo ve) .

Textbooks

Coates, D.F., (1967) "Rock Mechanics Principles" , Canadi an


Dept . Energy, Mines and Resources, Monograph 874.
Hoek, E., and Bray J. (1974) ItRock Slope Engine erin g' !, (In s t.
1

of Min and Metal, London) .

*Can be ordered through ISRM , Laboratorio Naciona1 de Engen hari a Civil,


Avenida de Brazil, Lisbon, Portugal.
Introduction 13

Jaeger, J.C., and Cook, N.G. W. (1969) "Fundamentals of Rock


Mechanics", (Methuen).
Krynine, D., and Judd, W. (19 59) "Pri nc iples of Engi neering
Geology and Geotechnics", (McGraw Hill).
Ob e rt, L., and Duvall, W., (1967) "Rock Mechanics and the
Design of Structures in Rock", (Wi l ey).
Scott, R.F., (1963) "Principles of Soil Me ch anics", (Addi s o n
Wesley) .
2
rock classification

While this book primarily concerns the discontinuous rocks, it


is necessary to see this rock class in context and accordingly the
question of rock classification in general will be explored. The
object of rock investigations and measurements is to make judgments
about the rock as a prelude to some action. The properties used to
classify the rock will vary according to the designer's purposes and
may include various subsets of: shear strength; flexural strength;
tensile strength; elasticity; permanent deformability; creep-rate;
water flow and water storage properties; in-situ stress; drillability;
fragmentation characteristics; and sometimes density, thermal expan-
sion, mineralogy, and color.

THE NATURE OF ROCK

One can not assign rock properties to a design calculation with


the same degree of certainty as with some other types of engineering
materials. The reason is that there is rarely a wholly dependable
large sample of the total population available from which test results
can be extracted. The application of principles of structural geology
makes the sampling problem solvable. But we must realize that most
of the volume of rock of immediate concern is hidden and inaccessible
and, unfortunately, what we do see is rarely representative of what we
don't. It is almost a law of geological engineering that the hidden,
Rock Classification 15

mantled material is the weakest and potentially most troublesome ; only


the sandstone layers will cropout in a formation composed of sandstone
and shale; only the flow rocks will form ledges in a volcanic series
of basalts and pyroclastics. The granite will form a hill, but the
fault zone through it will form a valley.
Nor can the designer of a work in rock make u se of rock prop er -
ties with the same rigor as he might for other types of st ru ctural
and hydraulic computations, because rocks seldom lend themselves to
the usual sort of idealizing assumptions. First, most rock f o rmations
have directionality, such as bedding in sedimentary rocks, flow band-
ing in volcanic rocks, and foliation in metamorphic rocks, and are
consequently moderately to highly anisotropic. Then we fin d rock
responding differently to excavation according to the i n it ial state o f
stress , particularly in underground applications, and this is heavily
dependent on the stress history which will be known only occasiona lly.
Many rocks are semi -discontinuous on the hand specimen scale owi ng to
a network of fissures and flaws, and almost all rocks on the formation
scale are penetrated by surfaces of potential or real disco n t inuity.
At the depths reachable in deep mines, deep drill holes, and some
tunnels, some rocks are ductile, and very few rock s behave en t irely
elastically even at low pressures. Some rocks are chemical l y change-
able within the lifetime of an engineering work a n d even more show
great variability vertically and horizontally, due to different de g rees
of weathering. In the face of these difficulties, results of comp u t a-
tions are to be utilized with restraint, and controlled by o bserv a-
tions during construction. Fortunately, it is often sufficient for
engineering purposes to produce only a reasonable estimate of the
final behavior - an estimate that can be ar r i v ed at sati s f actor i ly by
rock classification.

ROCK SPECIMEN VERSUS ROCK-M ASS

In a discussion of rock classification, we must carefully dis-


tinguish characteristics of a specimen of rock from properties of a
body of rock in situ which, in the language of rock mechanics, we c a ll
the rock mass. The mass is comprised of the rock, its network of
discontinuities and its we a t hering profile. The behavior of the roc k
16 Rock Classification

mass reflects all of these components as well as water and stress


regimes, strength, deformability, and permeability, which may be
largely unrelated to material properties.
Classification of the entire realm of rock masses for the to-
tality of applications would demand an unwieldy number of independent
factors because different pursuits require different parameters. In
assessing the suitability of facing stone, aggregates, embankment
materials, and other rock products, we need rock specimen attributes
describing durability, strength, thermal expansion, shrinkage, swell,
absorption, and specific gravity. Rock mass characteristics affect
items related to the cost of production. In regard to excavations,
both specimen and mass characteristics are essential, the former
affecting drillability and durability and the latter being basic to
stability while also influencing excavatability. The essential
factors for foundations, particularly for hydraulic structures, are
those descriptive of rock mass deformability, stability, and per-
meability which derive principally from the discontinuities, (although
rock specimen .characteristics may sometimes control the design, as
for example in non-durable, fissured, weathered, or permeable rocks).
First, we will examine classification of rock specimens then
the weathering profiles and systems of discontinuities and finally
the classification problem for rock masses.

PETROLOGIC CLASSIFICATION OF ROCK SPECIMENS

Geological methods of classifying rock specimens are based on


a number of different criteria, which can be studied in Williams,
Turner and Gilbert (1958). We will explore the wisdom of using geo-
logical rock names and petrological descriptions for engineering
purposes.
A description of a rock's texture and fabric affords a basis for
understanding its mechanical properties, which are closely related to
interparticle bonding, interlocking and imperfections. The crystalline
rocks (figure 2-1a) have tightly interlocked particle arrangements
sometimes impaired by micro fissures within and between crystals.
Coarse grained crystalline rocks tend to be weaker and less stiff
than fine grained or aphanitic crystalline rocks. Foliation, the
Rock ClassificatIOn 17

(a)

(b)

(c)

Figure 2· 1. (a) High ly inter locked, crysta ll ine texture of a Mesozoic quartzite; (courtesy f Prof R. WenkL
25.5X . (b) Highly anisotropic crystal l ine texture with oriented fissures (fracture 1.ledvage) III chlorite
schist- Ho mestake M ine, S.D.; (courtesy of Dr W Chinn). 40X (c) Porous, clastic texture ' eolian sand,
stone from Olduvai gorge, Tanzan ia, consisting of poor ly sorted rock fragments and grains, some coated
with clay. Cavities occur in altered nepheline grains (N); (courtesy of Prof. R. Hay). 136K
18 Rock Classification

most predominant fabric element of metamorphic rocks, causes strong


anisotropy and surfaces of weakness within the scale of the specimen
(figu re 2-1b). Foliation is particularly pronounced when formed by
coplanar platy minerals like mica. In clastic rocks (figure 2-1c),
grain size has far less influence on mechanical properties than the
nature, strength and durability of the binder or cement. Properties
of cemented and compacted varieties of shale, for example, can be as
different mechanically as soil and rock. Bedding is the most im-
portant structural feature of sedimentary rock on the specimen scale,
as well as in the rock mass; it creates anisotropy in all properties.
Since geological names for rocks are intended to classify rocks
according to differing modes of origin, one may wonder if they are
meaningful for geological engineering practice. In the igneous rock
group, the genetic division between intrusive and extrusive rocks is
meaningful in terms of engineering attributes since it concerns the
depth of formation. Features derived from the surface environment --
vugs, amygdules, and flow structures -- partly determine the mechani-
cal properties of the volcanic flow rocks. The plutonic rocks, on
the other hand, present quite different aspects linked to their
formation at depths of perhaps 30 miles where the pressure approxi-
2
mates 150,000 psi (1000 MN/m ). For example, plutonic rocks such as
granite may possess large horizontal stress and fissures from un-
loading and a strong inclination towards chemical weathering.
Dynamically metamorphosed rocks (as opposed to products of thermal
metamorphism) contain miniature fold and fault structures and minerals
oriented during growth under deviatoric stress. The various genetic
processes responsible for the sedimentary rocks also produce distinct
assemblages of properties linked to the mode of origin, -- although
on the specimen scale the mechanical properties are more directly
related to textural and mineralogic considerations independent of
origin.
Mineralogic classifications form the basis for the actual rock
names in the igneous and metamorphic rocks and to some extent in the
sedimentary rocks. The mineral composition of crystalline rocks is
not vital to a classification of mechanical properties and consequent-
ly many of the rock distinctions important to petrologists are useless
Rock Classification 19

for engineering purposes; for example we usually don't care whether


a rock is classified as a granodiorite or diorite or tonalite.
However, the accessory minerals may vary from one species to another
and these, more than the proportions of quartz and feldspar may
affect engineering response. Pyrrhotite (Martna, 1970), possibly
pyrite, iron-rich micas, nepheline, leucite and nontronite have been
identified as instrumental in deterioration of originally solid rocks
quarried for aggregate and building stone. Minerals containing vugs
filled with carbon dioxide can lower the pH of the groundwater, con-
tributing to rapid weathering as at Bergeforsen Dam (Aastrop and
Sallstrom, 1964). Any of the sheet silicate minerals, e.g. mica,
chlorite, talc, and serpentine, introduce low shear strength,
especially if in coplanar orientation; mica schist, serpentine and
talc schist can be hazardous rocks in foundations and excavations.
Glass, and secondary minerals zeolite and opal, can promote chemical
reactions with cement even when present in small quantities in rock
aggregate. Crystalline sedimentary rocks include some varieties
largely or partially composed of weak, soluble, or non-durable grains,
e.g. clays of the montmorillonite group, gypsum, halite, sylvite,
we ak shales, coal, chalk and chert.
In summary, though the science of petrology has evolved accord-
ing to the needs of classical geology, its refined terminology and
class distinctions are frequently meaningful for engineering work.
Moreover, as geologists are familiar with it, and its rock classes
are generally mappable, the geological nomenclature, especially when
accompanied by textural descriptions and mineralogic details, is
the most appropriate rock material classification for engineering
purposes. The complete classification of the rock material must of
course describe the state of weathering, the durability, and the
degree of fissuring.

ROCK VERSUS SOIL AN D WEATHERED ROCK

The most vital distinction to be recognized is between rock,


weathered rock and inherently soil-like rock. The distinction is
essential for all engineering work in rock, and yet it is not an
Goodman-Geo log ical Eng.-3
20 Rock Classification

elementary proposition. The fundamental precept is that to be rock,


the material must be strong and durable. It is solid when first
encountered and can not be softened, disaggregated, or' easily weak-
ened by accelerated weathering. Furthermore, it does not swell or
shrink appreciably upon soaking. These requirements are pragmatic but
do not coincide with geological nomenclature, in which a rock is
defined as "any consolidated or coherent and relati vely hard,
naturally formed mass of mineral matter".*
Table 2-1 was based upon one devised by Karl Terzaghi for
students in his engineering geology class** to distinguish between
rock, weathered rock, and soil-like rocks. One may apply the term
"solid rock", according to Terzaghi only if a rock is solid with
a ringing sound when struck by a hammer and remains solid throughout
weathering tests and soaking. Moderately soluble varieties, such as
limestone and dolomite, will still be classified as solid rock, but
greatly soluble rocks such as salt and gypsum will not survive a
reasonable weathering test intact. Rocks which are originally solid
but break up into small, hard pieces with a clean surface on weather-
ing are termed fissured or crushed unaltered rocks, whereas if the
rock disaggregates or yields greasy surfaces, it is an unstable or
slightly decomposed rock. If such a rock exhibits perceptible
volume change upon soaking, Terzaghi thought "rock" would be a
dangerous misnomer; he preferred to designate swelling materials as
"intermediate between rock and clay, rock characteristics predominat-

* Dictionary of Geological Terms" Dolphin Reference Book C36D. The


above is the ordinary usage but this dictionary gives as a strict
definition "any naturally formed aggregate or mass of mineral
matter whether or not coherent, constituting an essential and
appreciable part of the earth's crust". The word consolidated in
the first definition is troublesome to engineers familiar with
soil consolidation theory which refers to the expulsion of water
from the voids of a soil under pressure. The geological usage
means firm.

** Table 2-1 is based upon one devised by Karl Terzaghi and distributed
to students in his course on engineering geology at Harvard University
in the 1950's. A copy revised shortly before his death was generously
supplied by Dr. Ruth Terzaghi. A somewhat similar approach is used by
the National Institute for Road Research, South Africa, as published
by Weinert (1964); see Fookes, Dearman, and Franklin (1971).
TABLE 2-1
EF FECTS OF SATURATION ON ROCKS AND ROCK-LIKE MATERIALS
Terzaghi's Guides for Distingu ishing Rock, Weathered Rock, and Soil *

Afte r r e peate d drying, Vo l u me change


In i mme r sin g, and sh a king , produced by
original or up on prol onged s a t ura ting dri ed Group
state expo sure to the f ragments wi t h
atmosphe re water

unchanged a) so l i d rock

bre aks up i nt o small b) f in e ly fi ssure d


h ar d pie ces wit h or crush e d
clean surf a c es un al te red r o ck

break s up into small c ) s li ghtl y de -


frag me n ts with "greasy " compos e d f issure d
s ur fa ces owi ng to the rock
pre se nce of fi ne - impe rc e p t i b l e
grain e d weat he r i n g
Solid pro duct s
wit h
ringing breaks up i n to ind i- d) sandst one or
sound v idual sand or s ilt muds t on e wit h
when part icl es un s t ab le c e me nt
struck
with a
hammer break s up i n to small e) i nterme d iat e
angular fr a gments with- between r ock a nd
out a ny indi catio n of c l ay, rock c h arac
ch e mical a lte rat ion t eristics dom i nan t

gradual ly t ransf ormed f) i nt erme di ate


into a suspen sion of measurable be twee n rock a n d
so i l p art icles c l a y , clay c har ac
teristics dominant
gra du a l l y t ransformed g ) thoroughly
i nto a s us pen s i on of decompose d r o ck
Solid c l a y part i c le s a n d a
with sedime n t consis ti ng of
du l l angula r rock f ragmen ts
sound
when comp l et ely t ran sf orme d impe rcept ibl e h ) c l a y , silt , a nd
struck into a s usp e nsion and / to ver y fin e san d in
with a o r a loos e sed i ment impor ta n t d r y o r a very com
hammer pa ct ed con d i tion

* From Professo r Kar l Terzagh i' s 00urs e note s fo r Engine e ring Geology
a t Harv a rd Uni versi ty; i n cl ude d wi t h k ind p ermission o f Dr . Ru th
Terzagh i (w i th minor e di t ori al c h ang e s) a nd in cludi ng rev isions made
b y Kar l Terzagh i sho rtly b e f ore his death.
22 Rock Classification

ing". Materials that are not solid with a ringing sound when struck
by a hammer when first encountered should not be referred to as rock
at all, according to this scheme. Many sedimentary rocks would
accordingly be termed "soil-like rocks II in maps and reports, and the
resulting impression would be correct for the engineer.
Geological investigations must correctly diagnose a specific
soil-like condition as either inherent or localized. Weathering, and
hydrothermal alteration -- the first usually intensifying towards the
surface and the other, with depth or laterally -- may produce spotty
and variable degrees of localized softness. In contrast, some
sedimentary rocks are inherently soft either through incomplete
cementation, intense fissuring, or regional alterations; neither
"dental work" nor outright "extraction" can improve the rock condi-
tions in this case.
The distinction between rock and soil is especially important as
regards specifications for excavation contracts. So many legal
controversies have revolved about this point that agencies such as
the U.S. Bureau of Reclamation have been forced to adopt almost
comically detailed wording for contracts, as in Table 2-2. The main
ideas are that the material to be excavated is rock only if it is
both in place, (or of large mass) and solid. If it is too risky to
attempt a classification, the excavation receives one name--unclassified
excavation -- and one price throughout. This can happen in deeply
weathered materials, with their extreme variability and gradational
qualities, in soil-like soft rocks, in bedded rocks alternating in
hardness, and in very dense or cemented soils.

WEATHERING

Closely related to the question of differentiating soil from


rock, is evaluation of the degree of weathering of the rock material.
The importance of the subject is suggested by a voluminous literature,
a selection of which is included in the list of references. Rocks
respond to prolonged weathering in many ways. The granitic rocks
become cracked and then decayed by the carbonic acid developed as
rain water filters through the soil; this reagent attacks the feld-
spars and dark minerals releasing soluble salts of K, Mg, Fe, and Na,
Rock Classification 23

TABLE 2-2

Classification of Excavation
According to U.S. Bureau of Reclamation
Contract Specifications

"Except as o therwise pro v ided in these specificat ions, ma te r ial


exca vated will be measured and classified in e xcavation, to the lines
shown on the drawings or as prov ided in these specificat ion s, an d
will be classified for payment as follows:

Rock Excavation. For purposes of classification of excavat ion,


rock is defined as sound and solid masses, layers, or l e dges
of mineral matter in place and of such ha r dn ess and t extur e
that it:

(1) Cannot be effectively loosene d or broken down b y


ripping in a single pass with a la te model tra c t or-
mounted hydraulic ripper equipped with o ne dig g i ng p oi nt
of standard manufacturer's design adequately s i z ed for
use with and propelled by a crawler-type tract o r ra te d
between 210- and 240-net flywheel horsepower, ope rat i ng
in low gear, or

(2) In areas where it is imp r actic a b le to clas s i f y by


use of the ripper described above, rock e x cava tion is
defined as sound materi a l of su c h hardness a n d t e x t ure
that it cannot be loosened or b ro ke n down by a 6 - pound
drifting pick. The drift~ng pi c k s h all be Class D,
Federal Specification GGG-H-506d, with handle no t le ss
than 34 inches in length.

All boulders or detached pieces of solid ro c k more th a n 1 c ub ic


yard in volume will be classified as ro ck e x cavat ion.

Common Excavation. Common excava~ion i nc lude s a ll ma teri a l


other than rock excavation. All boulders or de ta c he d pi ece s
of solid rock less than 1 cubic yard in vol u me will b e class i-
fied as common excavation."

as well as free silica which may be transported out of t h e weat her i ng


environment , and detrital clay and resistant q ua rtz gr ain s whi ch
usually remain. The rock is gradually tra n sfo r me d into a "saprolyt e fl ;

figure 2-2a, which resembles rock but has the st ren g t h o f a dense
soil. Vargas (1953), Ruxton and Berry (195 7 ), Lumb ( 19 6 2) , Dee r e
and Patton (1971), and others have described the tran si t ional state s
24 Rock Classification

(a) (b)

Figure 2-2. (a) Decomposed granite. Former joint blocks contain hard " core stones"
in their centers while relict joints are now sandy clay seams or partings in the thor-
oughly weathered rock (saprolyte). (b) The top-of-rock surface in soluble marble;
Columbia, California. The soil was removed by hydraulic monitors to obtain placer gold.

of granitic rocks and their properties. Basic igneous rocks follow


a similar sequence but tend to produce a residuum richer in clay. The
soluble rocks become enriched in impurities, often clayey, and develop
stable or unstable vugs according to their strength. Limestones are
often karstic whereas gypsum and halite are simply thinned or removed
altogether, the karstification inducing almost immediate collapse,
(Brune, 1965), (figure 2-2b). Intermediate weathering states con-
sisting of vuggy rock are less common in gypsum than in limestone
and dolomite rock. Anhydrite expands, relative to the initial solid
volume, as it is converted to gypsum by hydration (but relative to
the total volume of reactants it contracts).* Compacted shales and

* There is a difference of opinion on hazards of anhydrite. A


thesis by Sahores (1962) considered the engineering problems
implied by volume expansion to be overstated. Brune (1965)
on the other hand, described uplifts and explosions occurring
naturally in an area of West Texas underlain by anhydrite at
depth; moreover the anhydrite grades into a thinner, folded
gypsum layer updip and the uplifts occur directly over the locus
of anhydrite - gypsum interfingering leaving no doubt that con-
version of anhydrite to gypsum is responsible for these violent
phenomena.
Rock Classification 25

po o rly - ceme nted san dst ones -- t he so il -l ike r ock s -- disaggregate


an d retu rn to sedimen t i n re s ponse to weathering, and montmorill on itic
v ari etie s swel l . In genera l , as t h e non-soluble rocks pass through
intermediate stages of we at h eri ng they g ain po rosity and deformability,
l ose stre ngth a nd e las t ici t y , a n d become first more and then less
p erme abl e; (s ee f or example Iliev's (19 66 ) di scussion of property
cban ge s in weathered monzo n i te ), To classify t he materials and
a tt rib utes of t he wea t h er ed z ones, one must consider t wo i n dependent
c r it eria: fi r st it is ne ces sary t o distingu i s h differing degrees of
wea t heri ng of t h e r ock itself; t h e n th i s d i sti nct ion must be super-
i mpose d on a class ifi ca tion of dif f ering s t yles and arrangements of
t he weathe rin g produ c ts.

The Degree of Weathering

Ap p r ai sal of th e deg re e of we a ther i n g ac tu a l l y attained b y a


p a rt icul ar specimen of ro c k mater ia l is bas ic to any meaningful
c lassifi cat ion of ro c k masses wit h in t h e weat hered zone, wh ich in the
tr op ics a nd in par t ic ula rly susce pt ible rocks s uch as granite, may
exten d more t han 300 feet b elow t h e g ro un d sur f a ce. No single index
de ri ve d f r om Si mpl e field ob servat i on s o r labo r at ory tests can be
e xpe c ted t o a p p ly a pp r opriately f o r all ma ter ia ls in the vast range
of weather in g pro ducts de ri vable from intermediate stage s of decom-
pos i t i o n o f ro ck , Several ap p roac h es useful in particular rock types
a r e off ered as e xamp les to b e emulated i n pri nciple or detail as the
cas e wa rr ants.
Lumb (196 2 , 1965) disc us sed co rrelati on between soil and rock
p rop e r ties in g ran ites of Hon g Kong ordered b y a mineralogic weather-
in g inde x X , Lumb's inde x , ap pro pri ate f o r quartz b e a ring granitic
d
rock s i n which the felds p a rs are attacke d du r ing the decomposition
p roce ss , is asse ss e d by han d lens ex ami na ti o n of weathered and fresh
rock to dete r mine the percentages of felds pa r and quar t z as follows :

eN q - Nqo ) / (1 - Nqo )
Nq is the weig h t r a tio of quar t z to quartz + feldsp ar in the weathered
speci men , an d Nqo i s t h e c or re sp ondi n g r atio i n th e intact , un-
weathered s peci me n . N is of th e o rder of 1/3 for a fresh granite
qo
a nd increase s t owa r d 1 as t h e weathe ri ng progresses. Thus the index
26 Rock Classification

varies over the range 0 to 1 with an increasing degree of weathering.


Ege (1968) used a similar approach as one of four classification
indices for granitic rocks at the Nevada Test Site. The degree of
weathering is expressed by estimating the percent of altered minerals
in the rock, without reference to an unweathered standard. The
rock is classed as unweathered, slightly weathered, moderately
weathered, or severely weathered respectively as the percent of
altered minerals falls within the classes 0-10%, 10-25%, 25-75%, and
75-100%. The degree of weathering can also be classified on less
formal divisions as in the example by Kiersch and Treasher (1955) for
granodiorite at Folsom dam, California where: fresh rock was totally
unaltered; slightly weathered rock showed slight fissuring in the
feldspars and bleaching of their original color; moderately weathered
rock showed more intense bleaching and fissuring in the feldspars,
bleaching of the biotite, limonite appearing as specks and coatings
of other minerals and slight rounding of quartz grains; and highly
weathered rock showed strongly bleached biotite, the feldspars highly
fractured and bleached, the quartz grains highly rounded, and
limonite common as an accessory; further, the highly weathered rock
could be scratched readily with a steel nail. This simple classifi-
cation could be mapped and was successfully correlated with variations
in resitivities, seismic velocities, drilling rates with diamond and
percussion drilling, blasting patterns and powder factors, rippability,
grout takes, and suitability of stone for rock fill and rip-rap.
Iliev (1966) introduced an index K based upon the reduction of
longitudinal wave velocity with weathering.*

vo - Vw
K (1 )
V
o

The subscripts 0 and w identify the unweathered and weathered states.


Like Lumb's index, this one goes from 0 to 1 as weathering progresses.

* Such an index can be applied in the laboratory or in the field; in


the latter case characteristics other than specimen properties
are involved and classification by the application of this simple
parameter can be wrong.
Rock Classification 27

Hamrol (1961) proposed a simple measurement of apparent porosity


by the water content of a rock (dry weight basis) as an index of
degree of weathering after quick immersion. The water content is
determined after oven drying at 105 degrees centigrade. Lumb (1962),
Pender (1971) and others have shown that porosity increases with
weathering (see figure 4-16); since some engineering properties of
rocks are directly associated with porosity or indirectly sensitive
to its changes (Griffith, 1937), it is not surprising that Hamrol's
index has met with success (Serafim, 1964) in recognizing rock grade
boundaries within a single rock type at a single engineering site and
in extrapolating results of field tests from one part of a foundation
to another. There has been little quantitative work on the changes
in properties of joints resulting from weathering.

The Profile of Weathering

Most engineering projects involve rock work in various levels


within the weathered zone, which may extend as deeply as 100 meters
below the surface. The outstanding feature of the weathering zone is
extreme variability of rock quality, both laterally and vertically,
(figure 2-2a); rocks of various degrees of weathering grade into one
another insensibly. Classification of the weathered rock mass can
be meaningful if described in terms of percentages of various weather-
ing products at any given level (weathering horizon). Deere and
Patton (1971) reviewed the weathering profiles of different rock
types and suggested standard terminology based upon the approach used
by Ruxton and Berry (1957) for granite soils of Hong Kong. These
papers, as well as the work of Fookes and Horswill (1970) Spears and
Taylor (1972), and others listed in the references should be consulted.

Durability

The discussion of weathered rock has considered only observed or


measured attributes of a present sample. What will the properties
be some years later, in response to construction and service? The
question of durability and its inverse, weatherability, is only
beginning to be answered by testing techniques and comparative data
meager data in view of the variety of engineering requirements. Some
28 Rock Classification

of the minerals suspected of contributing to weatherability in rocks


were listed earlier; now, we will consider a simple index test.

Franklin's slake durability test. Fookes, Dearman and Franklin


(1971), and Franklin (1970) developed a durability test consisting
of a standardized measurement of the weight loss of rock lumps when
repeatedly rotated through a water-air interface. Ten lumps of 40
to 60 grams each are oven dried and weighed, and then placed in a
standard test drum (figure 2-3a) whose circumferential wall is
constructed of sieve mesh (2 mm opening). The drum is rotated at

(a)
(b)

Figure 2-3. (a) Slake durabil ity appara-


tus (courtesy Soil Mechanics Equipment
Co., Glen Ellyn, Illinois). (b) Franklin
Point Load Testing Device (courtesy Soil
Mechanics Equipment Co., Glen Ellyn,
Illinois).
Rock Classi f ica tion 29

20 revolutions per minute for ten mi n u tes. The s l ow speed r e du ces


me c hanical wear effect in the agi t ating pro ce ss. The dry we igh t
reta in ed a f ter t he we atherin g cycle, exp re s s ed as a p e rcen t age o f
the original wei gh t, i s r ep o rted as t h e S l ake Durabil it y I ndex (I d ) .
Gamb l e (1 9 7 1 ), wh o ev alu a te d t h is i ndex in r el a tio n to o t h er dur-
ab i Ii ty and abrasi on tests used for aggre gates , f oun d t he sl ake
durabi l ity to be f ar gentle r a n d bet te r abl e to c op e wi th t h e large
range of durab i li ty respon se offere d by ro cks (he pre f err e d a
modified durabili ty i nde x b ased on 2 c y c les of rot atio n and dryin g ) .
Tests such as the Los An gele s ab ra si on te s t * a re mo r e sen si ti ve to
slight variations in dur a b i lity among ro cks to be considered for
aggregate.
To assess t h e weatherabi l i t y o f a ro ck , it i s me an ingful to
attempt t o simu late a project' s a nti cipate d we atherin g e n v ironment
at an accelerated rate. The pitfall is that un le ss o ne is ab l e to
incorporate all pert in e n t factors i n t he l aborato ry s i mulat ion, the
results will be diffi c u lt to i nterpret. The a d vant a ge of a stan-
dardized te s t , such as the o n e des c ri bed , i s that e xp er ien c e g a ined
i n assorted p r oj ects wi ll ev ent ually be g rouped i n a useful f orma t
for f u t u re r efere nc e.
Slaking of claystones and s h a l es can be c au sed by swe ll in g of
clays. Tb e slak e durabi lit y te s t is n o t su i t ab le fo r swe l lin g
materials as the lumps tend to bui ld p r ot e ctive c l a y co a t ings.
Dur ab i lity p roble ms a sso ciated wi th e xpans ive c l a y mi n era l s can b e
predicted by stan d a r d metho ds of testi n g fo r t he prese nc e of swe ll-
ing clays an d measu r e ment of swe l l p r essu re. A con so li dometer
especially suited to t hi s purp ose is the Ge oNor Swe lli ng app a rat us**
in which disc- shape d ro ck samples, o r pul v e r i ze d an d e lutria ted
samples, are pre c on so l idate d a n d then allowed t o s we ll un de r impo s e d
disp l acement c on st rai n ts ( Bje rrum, Brekke, et aI, 1963). F ree
swell of altered, ha rd rock samp les wa s meas u red witb s ufficient
precision very simp ly by Nasc ime nto ( 19 70) wi t h a jewe lled di al ga u ge .
The rock co r e s p eci me n st a n ds in a be a k er of wa t er on a p o i n t con-

* ASTM Standar d Me t hods of Test C5 35 -69 and C131-69 .

** So ld by the No r we g ian Geot echn i cal I n st i tute, Oslo.


30 Rock Classification

tact. In several altered granites and gneisses Nascimento monitored,


swelling began almost immediately after the water was added and
essentially terminated after two to five hours.
A meaningful and potentially rewarding area of inquiry into the
weatherability of rocks considers the changing content of dissolved
solids in water percolating steadily through rock specimens. The
Bernaix radial permeameter, discussed later in connection with
fissuring, is suited to this approach.

INDEX TESTS FOR THE QUALITY OF THE ROCK MATERIAL

Other simple laboratory tests or quick field measurements can


serve as quantitative indices of rock quality and degree of weathering
and as basic components of applied classifications. Deere and Miller
(1966) studied the use of the Schmidt hammer which can be carried in
the field. Defects on the surface against which the hammer is
activated can give low readings unrelated to the rock material
quality, a problem which can be avoided by exercising care in pre-
paring the test surface. A more revealing measurement is provided
by any strength test, especially one which demands a small sample and
which can be done routinely on a large number of specimens. The
point load test (figure 2-3b) introduced by Franklin (1970) and Broch
and Franklin (1972) is one such method. Tests are conducted by
squeezing pieces of rock drill core diametrically between standard
2
steel cones until rupture. The point load index I is PjD where
s
P is the load at rupture and D is the diameter. The results are
affected by the value of D, but size correction charts given by
Broch and Franklin allow all results to be expressed in terms of a
standard size (50 mm is recommended). The point load index I s ,50
correlates fairly well with the uniaxial compressive strength
divided by 24. The test can also be applied to irregular chunks,
approximating 50 mm in size.

FISSURED ROCKS

Small cracks and fissures may be contained in apparently intact


rock specimens. As opposed to pores, which are three dimensional
(a) (b)

(e) (d)
F igure 2-4. (a ) Scannmg electron
microscope photo 01 a series of pores In
Ber ea sandstone. fWeinbrandt and Fatt.
19691. 102 X. fb) Scanning electron
microscope photo o t another pore In
Berea sandstone ; (We inbrandt and FaIt,
1969). 1020X . (c) Polished section of
Mont Blanc granite. show i ng Intense
ftssunng . (courtesy of Drs. M. Panet
and C. T ourenQ, Lab. des Ilonts et
Chausees. PariS). l A X. (d) M,crophoto·
graph of a fiss ure fabriC reproduced In
a plastic fil m replica of a polished
diorite surface; (COurtesy of Dr. J.
Verd ier. Coy ne and Beli ier , Par ies). SOX.
fe) Fine fissures In granite, hi gh lighted
by dye and viewed in thin sect ion;
Huelgoal gr an ite . Bri ttany; (courtesy of
(e) Drs. B. Schnei der and J. Verd ier, Coyne
and Bel1ier, PariS) .
32 Rock Classification

(figures 2-4ab), fissures are short planar cracks of microscopic or


macroscopic size (figures 2-4 c,d,e). They occur as intercrystalline
cracks less than 1 micron to macroscopic (> Imm) in size, as inter-
grain cracks, and as multi-grain fractures. The presence of such
cracks as well as their significance in reducing the tensile strength
of brittle materials were appreciated by Hoek, Brace, McClintock and
Walsh, and others with regard to the Griffith theory of failure which
is based upon stress concentrations around such fissures. Habib and
Bernaix (1966) linked the degree of fissuring also with scale effects
in strength and deformation measurements, dispersion of results in
repeated measurements, and stress dependency of specimen permeability.
All of these effects were shown to be large in highly fissured rock
at low pressure and to disappear in non-fi$sured rocks, and in
fissured rocks at elevated pressure, within which the fissures have
closed. With respect to mechanical properties, it is the presence
of fissures more than any other aspect, wrote Habib and Bernaix, that
distinguishes rock from other solids. The French have held two
colloquia and an international symposium on rock fissures* and
correlations have emerged reinforcing Habib's belief that in fissured
rocks, mechanical properties are more closely dependent upon fissure
fabric than on mineral composition or texture.** It appears that
fissuring has a primary influence on static elastic modulus values,
hysteresis in load cycling, sound wave velocity, direct tensile
strength, resistivity and thermal conductivity of rock specimens.
The degree of fissuring in a rock reflects its history. Fissures

* The 1st and 2nd Colloquia on Fissuring of Rocks were published in


special numbers of "Revue de l' Industrie Minerale" respectively
15 May, 1968 and 15 July, 1969. The Nancy Symposium held October
1971 was published by ISRM.

** In introducing the 2nd Colloquium on fissuring in rocks, Pierre


Habib wrote: "One can now say that the properties of rocks are
essentially those of their fissures. The mineral matrix has only
a discreet role in the sense that if the rock is continuous it
is always over endowed either in rigidity or in strength. To
describe the fissuring of a rock is thus to define its present
state and the study of its mechanical behavior is first of all
the examination of the arrangement and development of fissures
up to destruction."
Rock Classification 33

can b e gene r a ted b y chemi cal weac herin g , unlo ading, heati ng and
c oo li ng , and most imp ortant l y b y lo c ali z ed cr ack ing accomp a ny ing
defo rm ation. Roc ks likely to be fo un d in a fissured state are:
v o l cani c fl ow ro c k s; fo liate d met amorph i cs, espe ci ally schistose
v a riet i e s; marble; pe gmatites a nd porp h y ri t ic or hypidiomorphic
gran it ic ro cks ; g rani te s exhibitin g cleavage; serpentine; chert and
sili c eous sha les; a nd qu a rtzi te.
The d e gree of fis s ur i ng s h o u l d b e a basic compon e n t of any rock
c l as sificat i o n s cheme. It c an be char acteriz e d through direct
observ at ion , or mo r e simply throu gh index tests.
A pol ishe d s u rface will of ten enable promin e n t fissures to be
observed with t h e n aked eye. A han d lens, binocular microscope, or
be s t of all, an ore micros c op e al lows st udy of fissure distribution
in t h e p oli shed sec t ion . Tourenq (1 9 6 9) di splayed f is sures in
p ol ish ed surf a ces by prep ari ng rep l i cas of the su rf ace adapti n g
tech n i ques f or e lectron mic ro scope spec i me ns des cr ibe d by Bradley
(19 5 4) an d J acq u et and Me ncarel l i ( 1959 ) , (fi g u r e 2 -4d ) . Schneider
(196 7 ) used dyes: Fusc hine ASA (basic), Victoria Bl u e (basic ) and
Auramine J ( aci d) t o s h ow fi ne fis s u re detail in thin sect io ns viewed
in tran s mitted ligh t with a pet ro g rap h ic mi croscope, (figure 2--4e).
S tu dy of fl u i d- fi l le d e pigen eti c i nclusions , e. g. in quartz, can
a llow r elat ive ev a l ua t ion of dif f erent directi o ns of fissuring,
(Ver d ier a n d Deicha, 19 71 ). Thes e pro c es se s c a n be te d ious and for
pract ica l en g ineer i ng work it may be mo r e app ropriate t o characterize
th e de g r ee of f is suring imp lic itly.
Pa r a meters of fi s suri n g c an be d e r i ved from pressure - volume
ch an g e cu r ves , sh ear a n d l ong it udi na l wav e velocity meas u rement s ]
comp a ri s on of d i r ect a nd in dire ct t en s i l e st reng ths, and the rat io of
perme abilitie s in t en s i on and compression. Mo rl ier (1 96 8), follow i n g
wo rk of Wa l sh (1965) , s ug gested c o mput ation of the vo lume of f iss ures
--f i ss ure p oro sity f rom t he sh ap e of t h e p res sure - vo l ume change
cu rve (p versus tV ). As s h ow n in fig ure 2-5 , this curve is concave
upw a rd , b e c omi n g a symptotic to a l i ne wh ose sl o pe is defined by the
compre s s ib i li ty of t h e r ock (k = 3( 1~2 V ») ' The i n itial fis sure
p o rosi ty nf (o ) is est imate d b y the value of ~v at intercept of t h e
a sy mptote. The con cavi ty of t he cu r ve c a n also be interprete d to
34 Rock Classification

original rock
volume = V

Figure 2-5. Compressibility of fissured 6V


rock.
V

yield a fissure shape distribution function ("fissure spectrum").


Tourenq, Fourmaintraux, and Denis (1971) offered a second
approach based upon a comparison of actual and theoretical wave pro-
pagation velocities. A crystalline rock composed of given percen-
tages of stated minerals has theoretical elastic properties close to
the weighted average of the elastic properties of the components.
Table 2-3 gives values for Young's Modulus, Poisson's ratio, and
longitudinal and transverse wave velocities for the common rock
forming minerals. If a rock is fissured, measured properties will
be lower than the theoretical values calculated from Table 2-3. The
degree of fissuring is expressed in terms of a quality index, IQ,
defined as the ratio of measured to calculated longitudinal wave
velocities.

V$I, measured
IQ x 100% (2)
V$I, calculated

Fissure porosity (n ) drives the quality index downward linearly


f
approximately 15 times as fast as normal porosity (n ) (spherical
p
pores). If one measures the total porosity n% (= np + n f ) as well
as IQ, figure 2-6 can be used to determine the value of n . This
f
Rock Classification 35

TAB LE 2-3

Av erage Elastjc Modulus and Velocity


of Longit udinal Waves for
Common Rock Forming Minerals *

Young 's Poisson' s Lo ngitudina l Transve r s e


Mo dulus Rat io Vel oc it y Veloci t y
E v V£ V
t
5
( 10 bar) km / s e c km / sec

quartz 9.6 0 . 08 6 .0 4. 1
ort h oc l ase 6. 7 0.27 5.7 3.3
plagi oc lase 8. 1 0 .28 6.3 3.5
biotite 7.0 0.25 5. 1 3.0
c a l cite 8. 1 0.30 6.7 3.4
mus c ov ite 7.9 0 .25 5.8 3.4
amphibole 12. 9 0 .28 7.2 4.0
pyrox ene 14 . 4 0.24 7.2 4 .2
olivine 20 .0 0.24 8.4 5.2
magneti t e 23. 1 0 ~ 26 7 .4 4.2

* From da ta of Ale xs androv , Be l ikov a n d Ryz ova , a re f e rence c ited


by Fo urma intraux and To ure n q ( 1 970).

fi gu re also shows t he re lati ve e ffect s of po res and f issures on t h e


rat io of meas ured to c alculated e lastic modu l u s v a l ues. A value of
n f ; 2 % reduce s t he i de a l elas tic mo du l us almost by half , whereas a
value of n ; 15% wou l d be re q u ired to a chi eve this eff e ct.
p
If bo t h t h e tr a ns vers e a nd l o ngit udina l wa ve velocities a r e
measured, t he de gre e of f i ss u ri ng c a n be derived fro m t hei r ratio .
Fi ssure d ro cks are not "idea l" mate rials and on e s hould not auto-
matically tr y to re p ort the r at i o Vt/ V£ i n t e r ms f a I' dy n amic
Poisson ls ratioll val ue. I ns t ead , To urenq et a l suggest Table 2- 4.
A third metho d of ev alu atin g t he de gree o f fis s uring i s bas e d
upon the ratio of s tren gt hs i n di r ect an d indirect tension tests,
( Tourenq a nd De n i s, 1970) . Dire ct t en s i o n te st s can be performed by
bonding mome nt - fr ee e nd pie ce s t o cy l indrica l r o ck spe cimens. Th e
Goodman- Geul oglc al Eng.---4
36 Rock Classification

~
o
u
~
>
"--.
iii
o
a.>
E
~
>
Figure 2-6. Relative effect
of fissures and pores on the
longitudinal wave velocity
and the modulus of elastic-
ity; (Tourenq, Fourmaintreau,
Emeas./Eeale. and Denis, 1971).

direct tensile strength, at, i.e. the average tensile stress at


failure, is greatly reduced by fissuring. An indirect tension test
(Brazilian test) can be obtained by compressing the opposite diameters
of a rock disc* causing a uniform state of tension across the vertical
diameter; the Brazilian tensile strength at B i.e., the tensile
,
stress at failure in a Brazilian test, is only slightly affected by
fissuring. Therefore, the ratio at / at,B is descriptive of the
degree of fissuring, as shown by Table 2-5 summarizing some data
presented by Tourenq and Denis. They recommend that the rock be
classed as: essentially non-fissured if at / at,~ > 0.8; very
fissured if at / at , B < 0.2.

Bernai x (1969) developed an index of fissuring intensity based


upon a radial permeability test. Water introduced under pressure in
the center of a thick walled cylinder of rock, figure 2-7a, produces
tangential tension stress as it flows divergently towards the outer
circumference. Conversely, convergent flow produced by directing
water from the outer circumference to the inner produces a tangential
compression. Assuming that the flow net is not altered by stress

* In Tourenq and Denis' tests, the length to diameter ratio of the


discs was unity.
Rock Classification 37

TABLE 2·4
Index to Degree of Fissuring
According to the Ratio of
Transverse to Longitudinal Wave Volocities

Description

< 0.6 non fissu red

0 .6 to 0.7 f is s u red

> 0. 7 very fiss ured

TABLE 2·5
Tension Test Index to Fissuring

Data from To u r e n q a n d Denis (19 70 ) .

Ro c k Fi ss ure length °t *
( mrn)
°t , B

Limest on e 0. 2 1.0
Limestone 1. 5 0.45

Gran i te 0.1 0 .93


Granite 0.3 0.7
Gran ite 1.3 0.50
Gr an i t e 2.5 0.34
Gr a nite 1 to 10 0 . 14
Gr a ni te 3 to 20 0.07

Basalt 0.1 0 .9
Basa l t 2 t o 10 0 . 15

* 0t = di re c t tensi on stre ng th; °t B Brazilian tensile


,
st r e n gt h .
38 Rock Classification

(a) (b)
'-4+-------0olitic
limestone
S=I

(/)
"-

~ 10-
8

150
100 12 10- 10
2~0~~0~~2~0~~~~~6~0~~80~~100
P (bars)

Figure 2-7. Radial permeability test; redrawn from Habib and Bernaix (1966).
(a) Radial permeameter (dimensions in mm). (b) Results for a porous and a fissured
rock; S = K_,1K50

dependency of permeability*, Bernaix derived expressions for the


water pressure gradient dp/dr as a function of r, added this as a
body force to the equilibrium equations and solved for or and 08 ;
the value of 08 on the inner wall of the cylinder (r = R ) is
1

± P
+ (3 )
2(1-v)

08 is tension for divergent flow where P is the water pressure


on the inner radius r = Rl ; (P(R 2 ) 0)
and
08 is compression for convergent flow, where P is the water
pressure on the outer radius r = R2 ; (P(R l ) 0),

* In actual fact the permeability K is a function of and since ° °


is a function of r, K depends on rand dp/dr can not be solved as
readily as in the paper. The problem is one of "coupled flow";
see Noorishad et al (1972).
Rock Classification 39

For the conditions of Bernaix's tests, Rl = 0.6 cm, R2 = 3 cm


and Poisson's ratio 'J = 0.2 giving 08 = 1.53 P. Figure 2-7b shows
the variation of permeability K for an oolitic limestone and for
Malpasset gneiss as P was varied between 100 bars outside to 1 bar
inside. The permeability K was computed according to the relationship

Q
K In (4)
2TT LP

wherein Q is the steady state flow rate (1 3 / t) and L is the length


of the cylinder. The permeability varied continuously over more than
3 orders of magnitude for the fissured gneiss whereas it remained
constant for the limestone. Bernaix recommends as an index of
fissuring to report the ratio (S) of permeability at 1 bar in diver-
gent flow (K_ ) to permeability at 50 bars in convergent flow (K ).
1 50
A sampling of results with a number of rocks are summarized in
Table 2-6.

TABLE 2-6
Radial Permeability Test Index to Fissuring

Rock Description of S
fissuring

Limestone porous, non fissured 1


Limestone porous, some fissuring 1.3
Limestone fissured 2.6
Grani te slightly microfissured 1.2
Quartzite microfissured 1.8
Mica schist fissured 4.8
Schist highly fissured 10 to 100
Malpasset gneiss highly fissured 7 to 200 right bank
1 to 50,000 left bank*

* The failure originated on the left bank.


40 Rock Classification

DISCONTINUITIES
Rock masses invariably include numerous surfaces of real or
potential discontinuity. Though somewhat artificial, we will dis-
tinguish between discontinuities and fissures. Obviously, there is
a continuous distribution of discontinuity surfaces according to
length. However, fissures within a specimen are included in a sample
of the specimen, thus subject to meaningful inquiry in the laboratory.
Laboratory techniques for samples of larger surfaces of weakness are
developing but the results are seldom exportable to the field without
additional field observations and tests. Thus fissures can be con-
sidered as rock specimen features whereas discontinuities cannot.
A single discontinuity includes two mating surfaces and a space,
or filling. The term "joint" which has come to be used in engineer-
ing contexts for all or part of the family of discontinuities in rock
masses, is unfortunately potentially confusing for structural
engineers, who use the term joints to describe points of connection
in steel structures; in geological usage the term is applied only to
penetrative, repetitive discontinuities without appreciable shear
displacement. However, as the term joint is entrenched in its
engineering geology context, it will be retained here.
On a geological basis, we can distinguish extension and shear
joints, bedding, banding, contacts, cleavage, schistocity, foliation,
sheared zones and faults, as discussed in standard works in geology,
for example Leet and Judson (1971) and Price (1966). With reference
to mechanical and flow properties of a discontinuous rock mass, we
require considerably more information than the geological identifica-
tion. In particular, load-deformation and strength properties of
discontinuities (see Chapter 5) make specific reference to a number
of controlling quantities, including the parameters of the peak and
residual shear strength-variation with normal stress, the initial
angle of dilatancy, the normal pressure required to prevent all
dilatancy, the maximum amount a joint can close, the peak and residual
shear displacement, the tensile strength, and the normal and shear
stiffness. Though no rational formula exists for extracting the
explicit joint parameters required for an analysis from field observa-
tions, careful and detailed descriptions of the many encounters in
Rock Classification 41

out c r ops, ex cav a ti ons , and in the core box a l low the who le syst e m of
disconti nu ities in a n y pro je ct to be d i vided int o a r el atively smal l
number of types. Us uall y the fiel d des cri pt ion wil l permit reason-
able estima tes fo r certain o f the q uan t ities me ntioned and the ir
contributi on t o the rock mass cha r ac t e ri st ics. The j o int system
prop e rt ies are de r ive d from ob se rvable featur e s o f: 1 ) the discon-
tinuity surfaces ; 2) t he bl ocks they de fi ne t hro ugh t hei r repetiti on
an d intersection; 3 ) an d the p ropertie s o f t h e spac e between t h e
bl o cks. Most of t h e se feat ur e s can b e de s crib e d deterministically ,
but are better exp ressed stat is ti ca l ly through dis t ribution curves
and numer i cal in di ces.
Prop erties of t he joi n t s urface s th ems elves i nclude orientation ,
exte n t, p lan ar it y , r ou gh ne ss a nd wav in ess, a n d t he strengt h of wall
rock a s peri ti es . J o in t sy stems us ual l y di s p l a y several preferred
o ri e n tat i ons and t hi s a s pec t a l o n e re s ult s in wholly d i fferent classes
of ro ck masses. Th is sub j e c t l ends i t se l f t o st at is t ical and graphi-
cal treatment by me an s of s te re og r aphic proje c tion , which wil l be
elabo ra t ed i n Chap ter 3. Join t "e xten t ", mean i ng the t otal area or
length, us ually can n o t b e me asured dire c t ly in t h e f ield; however it
can b e estimated occas iona lly from aeria l photos (see Chapter 4).
Ro u ghness a n d wavine s s, whi ch influ e nce t he f ri ct ion a n gl e s ] di l atancy,
an d peak s h ear (Pa t ton 1 96 6 , Goodman and Dubois, 1971) refer to the
local departures fr om p l ana rity a t s ma ll and large scales resp e ctively
(fi g ure 2-8) . Th e most c onve ni e n t roughness meas ure f or rock mechan-
ics p u rp oses i s i n terms of th e l oca l an gles with respect t o the mean
pl ane through all th e hills a nd val l ey s of a j oi nt surface . Most
joints can b e r e pre se nt e d q ui te wel l b y p l ane s. P a t ton (1966)
meas u re d r ough n es s a ng les f rom e dge v iews of s u rfaces in outcrop s and
cli f f faces ; t h e require d da ta can be ob t a in e d from photo g raphs as
di s cusse d i n Ch ap ter 4. Wh en the joi n t surface i t sel f is expos ed in
outcrop , repe a t ed measur emen ts o f d i p wi ll gene rate a scattered dis-
tr i bution of va lues, who se st a n da r d devi a t ion o r mean departure may
be an estimate of t he me an r oughne ss a ng le, a s discuss e d in Chapter
5. The rough n es s ang l es i ncr ease join t s h e ar strength at l ow normal
pres sure, but at hi gh e r no rma l p ressure, t h e strength o f t he wal l
ro ck asperit ies c on t rols t he s h e ar strengt h of the joints. A good
(a)

(c)

(d)

Figure 2-8. (a) A very rough fracture surface in granite; note the perfect mating of the joint blocks across
the tracture. (b) A very rough bedding plane in limestone; voids between beds resulted trom oversliding
of asperities accompanying mass movement downslope. (c) A rough bedding plane surface; the roughness
is tormed by ripple marks preserved from the depositional surface. Photo by Dennis Lachel, (courtesy of
the Corps of Engineers). (d) A rough joint surface; the roughness is created by the intersections of the
joint with bedding; erosion has accentuated the rel ief. The rock is an argillite. Photo by Dennis Lachel,
(courtesy of the Corps of Engineers). (e) A smooth surface whose mean plane parallels bedding. Slight
roughness exists because the surface wanders from one argillite bed to another. Photo by Dennis Lachel,
(courtesy of the Corps of Engineers). (f) A very smooth, wavy surface of discontinuity undercut in the
spillway excavation of figure 1-6. This is probably an ancient shearing surface.
Rock Classif ication 43

(e) (f)

i nde x of t h e stren gth of the wal l rock is provided by Schmidt


hammer re adings o n the j oint s urfaces . The scrat ch hardness va l ue
o f the wa l l rock is also indic a ti ve of its st rengt h. An appro priat e
scrat ch h a rdne ss s ca le de v is ed b y Ege (1968) u s es grades of scratch-
a b i l i ty by a stee l nai l a nd t he fin gernai l. Very hard rock sur f a c e s
canno t be sc ratch ed b y a s t ee l n a il; hard corre sponds to a fa in t
scratc h impre s sion wh i le i n terme d iat e ha r dness means a distinc t
s c r a t c h t r a ce c an be l e ft. If the material can be scratched by the
f i n ge rna i l , it is s o ft .
Th e s h a pe s a n d s i ze s of rock b locks f ormed b y t h e s ystem of
disc o n t in u ities depend e n ti r e ly on the ori e ntat ions and spacings of
the v ar i o us j oi n t s ets, (s e e fi gure 5-1). Join t blocks f ormed mainly
by one set of s ur fa c e s, as f o r e x a mpl e , layers i n bedded rock, may
be te r med " Tabu la r il . " Co l umnar" joint block s are f ormed by t wo s e t s
o f j o i n t s ( Bur t on , 1965) o r , as i n volcanic fl ows, by any number of
s ets p a ra l l e l to a gi ven axis. "Prisms II are f(~rmed b y thre e very
re gu l ar o rt ho go n al se t s. " We dge s" and " slivers" are forme d by
i rreg u la r joi n ti n g syst e ms , a s i n me t amorphic and graniti c rocks.
The si ze s o f th e joint bl ocks ob vi us ly de pend on t h e spacing of th e
va rious j o in t se t s . It i s us u al t o speak o f jo i nt spac i ng , rat h e r
t han b l o ck si ze , although bo t h numbers need to be appreciated. Table
44 Rock Classification

TABLE 2-1
Terminology for Spacing
of Discontinuities (after Watkins, 1910)

Descriptive Term Spacing

Bedding Jointing inches mm

thinly laminated fissured* < 0.24 < 6

laminated shattered 0 . 24 to 0.8 6 - 20

very thin very close 0.8 to 2.4 20 - 60

thin close 2.4 to 8 60 - 200

medium moderate 8 to 24 200 - 600

thick wide 24 to 80 600 - 2000

very thick very wide > 80 > 2000

* Called "comminuted" by Watkins.

2-7 after Watkins (1970) presents recommended terminology for various


spacings of bedding and jointing. Figure 2-9 shows rock masses with
various styles of jointing.
A description of the space between blocks includes the dimension
of opening between the rock walls (aperture) and the completeness of
filling with a description of the character and permanence of the
filling materials. Joints and faults may be "unfilled", "healed"

Figure 2-9. (opposite) (a) Sheet joints in granodiorite, Yosemite National Park, California. (b) Bedding
with tight joints in other directions. The bedding was undercut in excavation for a tunnel portal; rock
reinforcement was installed from the catwalk above the top of the excavation. Ruedi Dam, U.S. Bureau
of Reclamation. Photo by Dr. F. Heuze. (c) Closely jointed argillite with prismatic blocks; near Chivor
Dam, Colombia . (d) Bedding and two joint sets defining tabular and prismatic blocks of relatively small
size and very much larger wedge shaped blocks; near Libby Dam, Montana. Photo by Dennis Lachel,
(courtesy of the Corps of Engineers). (e) A metamorphic rock mass with three regularly spaced, tightly
closed joint sets and well developed schistocity. One well developed joint set (J1) forms the moderately
rough surface inclined towards the observer. The traces of two additional joint sets and of the schistocity
form three sets of parallel lines across the surface of J1. Near Don Pedro Dam, California. (f) Regular
columnar jointing, and completely broken, irregu larly jointed volcanic flow rock (Iatite). The columns
formed from shrinkage during cool ing, with their long axes perpendicular to the isotherms. The broken
rock represents columns that rode along on still molten material underneath. l\lear Tullock Dam, California.
46 Rock Classification

TABLE 2-8
Classes of Fau It Gouge Materials
(after Brekke and Howard, 1973)

Dominant Material Potential Behavior of Gouge Material in Tunnels

in Gouge At Tunnel Face Later

Swelling clay Free swell, sloughing Swelling pressure and


squeeze against support
or lining; free swell
with down-fall or wash-in
if lining inadequate.

Inactive clay Slaking and sloughing Squeeze on supports or


caused by squeeze; lining where unprotected.
heavy squeeze under Slaking and sloughing
extreme conditions. due to environmental
changes.

Chlorite, talc, Ravelling Heavy loads may develop


graphite, due to low strength,
serpentine particularly when wet.

Crushed roc.k Ravelling; Loosening loads on lining;


fragments or standup time may be running and ravelling if
sand-like gouge extremely short. unconfined.

Porous or flaky Favorable condition May dissolve, leading to


calcite, gypsum instability of rock mass.

(cemented) or "filled". Brekke and Howard (1973) distinguish the


five classes of fault fillings (gouge materials) listed in Table 2-8
in discussing the influence of fault gouge on tunnel stability.
Swelling clay fillings are potentially the most troublesome class.
Other clay fillings, and the sheet-silicate minerals chlorite, mica,
talc, serpentine and graphite, can introduce extremely low shear
strength, particularly if the thickness of filling is greater than
the roughness amplitude (Goodman, 1969). The clay: quartz ratio of
the filling has been demonstrated to be a sensitive parameter of
joint shear strength in British coal measure rocks (Taylor, 1973).
Rock Classification 47

Crushed rock fillings, and incomplete quartz or calcite fillings offer


potentially high permeability; they can erode and undermine the
adjacent rock in exceptional cases. This is particularly troublesome
in unlined water tunnels.

CONTINUOUS AND DISCONTINUOUS ROCK MASSES

From a geological point of view, rock masses are divided into


field rock units, e.g. formations, members, and zones. When such
units are defined so as to lump similar lithologic units or groups,
they are useful engineering divisions, mainly because they are
coherent, mappable entities. From an engineering point of view, it
makes sense to search for additional means of classification, motiva-
ted by pragmatic interests along functional lines. An enormous
literature reflects the timeliness, if not the frustrations, of this
pursuit.
A first order division can be made on the basis of the degree
of importance of discontinuities. If a rock formation is described
adequately by the rock material alone, it may be called a continuous
rock mass. More often, in connection with civil engineering projects,
the mass behavior is controlled by the discontinuities and the rock
substance description is almost irrelevant; such rocks should be
termed discontinuous. A formal classification can be made on the
basis of the relative importance of the system of discontinuities on
the significant properties; one such measure might be the ratio of
formation deformability to rock deformability as given in Table 2-9;
(or the ratio of shear strengths or permeabilities could be used if
more appropriate and if they could be conveniently measured.) In
jointed hard rocks, such as slate, gneiss, granite, quartzite, and
marble, the rock is so strong, so indeformable, and so impermeable,
that only the discontinuities contribute effectively to displacements,
to stability problems, and to water flow. Such rocks should be
treated as discontinuous. Unlike granular soil which is a discontinu-
urn that can be approximated by a homogeneous continuum for many
purposes; the discontinuous rock mass has very low porosity and any
appreciable deformation requires a dramatic increase in open space
through dilatancy. A "soft" rock on the other hand may belong to
48 Rock Classification

TABLE 2-9
Classification of Rock Mass Continuity

Name Ematerial IE field * Field Typical materials


fracture
porosity

Continuous 1 to 1.3 < 1% Some granites, massive


sandstones, and massive
limestones; many Tertiary
and Quarternary sediments
and argillaceous sedi-
ments of all ages.

Intermediate 1.3 to 2 < 1% Many granitic and meta-


morphic rocks; iron
oxide or calcite-cemented
sandstones.

Discontinuous > 2 1% to 5% Highly jointed granitic


and metamorphic rocks,
quartzites and silica-
cemented sandstones;
volcanic flow rocks.

Loosened > 5 Voids and Rock slides; fault zones.


cavities
between join t
blocks which
are in edge to
face contact.

* Efield means here "modulus of elasticity" as determined by using


total deformation measurements from a field test in a formula
based upon the theory of elasticity.

the class of continuous rocks, for its intrinsic rock deformability,


strength, or permeability ("matrix permeability") may override the
respective contributions from the system of discontinuities. Most
sediments of Tertiary or Quaternary age, many older argillaceous
sediments, soluble evaporites, chalk, and friable sandstones can be
considered continuous. Weathered rocks should be classified separate-
Rock Classification 49

ly through re fere nce to t he weat h eri n g profi l e as previousl y dis-


c us s e d. Some bard ro cks are a l mos t free of discont in ui ti es and can
b e cl a s sed as continuous, e. g. some granit e s, massive sands t ones ,
an d massi ve lime stone s.
The beh a vi or of a p artic u la r joint de pends not only on i t s own
p rope r t ie s b u t on t he initi a l s t ress a n d wa t er press u res a l ong i t.
Th e same fo r mations wi l l ch a n ge fr om discon ti nuous to continuous wi t h
i nc re asi ng de p t h. Sinc e a j oint is ve ry thin , it can be t reated as
esse n t ially t wo d imen sio n al ; its st at e of stres s therefore is i n
e q ui l ib rium wit h the s tre ss s tate of the adjoining rock block s . We
can no t us e the concept o f j oin t stre ss i n discussing a rock mass t hat
h as b een di s p l a c e d b y sli d i n g b e cause blo ck fo r ce s are t hen t rans-
fe r r ed pa rt l y t hro ugh e dge-to- f a ce contac t s across joints. A rock
ma ss with o p en j o i n t s a n d wit h block s i n e dge-to-face con t acts is as
differen t f rom a r o ck mass wi t h mati n g jOint surfaces as roc~ is
di ff ere nt fr om gr ave l ; su c h a r o ck mass t c h aracterized by open join ts ,
lo ad t rans f e r renc e al ong p oi n t a n d line contac t s , and in te rblock
c a v it i es, s h all b e c al le d l oos en e d.

ENGINEERING CLASSIFICATIONS OF ROCK MASSES

Tab l e 2- 10 give s r e f erences to some e n g ineeri ng classificat ion


systems f or ro ck ma t er ia l and r ock ma s ses. In con t inuous rocks I it
is s uf f icient to c lassif y t he rock mater i al alone and we can app l y
c l assification s of the t yp e p ropose d by Coat e s (1964), motivat e d by
i n t e r e st i n minin g problems u n der g round, or De e re and Mille r (1966) ,
motiv a t e d by an int e res t in s tanda rds for laborat ory t esting. All
of the ref ere n ce s to g e n e r a l pur p o se c lassi f icat ion given in Tabl e
2- 10 are def i c ien t i n omi tt i ng e xp l ic i t refere n ce t o mi cro -fissur i ng .
Ge ne r a l purp o se rock mas s classifications fo r d i scontinuous
ro c ks h a ve been p rop osed by n ume rous aut hors , a sampling of whom are
li st ed i n Table 2 -1 0.
Omitt i ng re fer en ce to t he we a th ering pro f i l e s, previously dis-
cus s ed, a f u nc t ional c l as si f i cat ion mu s t att emp t to overlay a class i -
f i c ation of d iscon t inui t i e s on o n e or more i n dice s of rock materia l
b e h a v ior. On e typica l s c heme, f or e xamp le, by F r a nklin, et a l (1971)
superimp ose s divisi o n s of fra c tu r e spaci ng and r ock strength (figure
50 Rock Classification

TABLE 2-10
References to Some Engineering
Classification Systems for Rock

Object For general purpose For a special purpose

Rock Coates (1964) Bergh-Christensen and Selmer-


Material Olsen (1970)
Coates and Parsons (1966) - resistance to blasting

Deere and Miller (1966) Selmer-Olsen and Blindheim


and Deere et al (1967) (1970)
- drillability
Underwood (1967)
- shales

Rock John (1962) Terzaghi (1946) - tunnels


Mass
Onodera (1970) Lauffer (1958) - tunnels

Iida et al (1970) Bieniawski (1974) - tunnels

Muller and Hoffman (1970) Barton et al (1975) - tunnels

Franklin et al (1971)* Kruse et al (1969)


- tunnel liner design

Ege (1968)
tunnels in granitic rocks

Obert and Duvall (1967)


- mining

Goodman and Duncan (1971)


- rock slopes

Caterpiller Tractor Co. (1966)


rippabili ty

Best applied to rippability classification.

2-10). The rock strength is to be determined either by unconfined


compressive strength tests or by the point load index, previously
discussed. This approach can be useful for organizing case experi-
ences.
A number of workers have considered the specific problem of rock
mass classification for tunnel excavation and supports. Barton,
Rock Classification 51

Uniaxi a I compressive strength, M P


1.25 5 12.5 50 100 200
Very Ex t rem ely

I
Weak Strong

Very
EH Thick

2~~----~~--~-+----+-~--~----~~-----r

VH Thick

01
E 0.6~--~--4-~~~--~~+-----~~----4-----~ C
U
H
o
0..
H Medium Vl
01 Q)
C c
u 0.2-+---4--4---~~~---=~~~~------4-----~ E
o 0..
0..
Vl 01
M Thin C
~ U
:3 U
gO.06~----~4-~--~~----+------+------+---~~
Q)
aJ
Lt 00
Very
L Thin

0.02 ~------~=-----t-----O--+---~----+---=-+-

VL VL

0.006~------~-----r------+------r------~--~"
VL L M H VH EH
0.03 0.1 0.3 1 3 10
Strength, Is ,MP
Figure 2-10. Franklin's rock
EH Extremely high L Low
quality classification; Franklin
VH Very high VL Very Low
H High EL Extremely Low et al. (1971). The strength Is
M Medium is the point load index .

Lien and Lunde (1975), for example, adjusted six parameters by means
of detailed study of 200 underground case histories, in proposing
a single numerical tunneling index -- Q.

RQD J J
(5)
r w
Q
r-
n
Ja SRF

RQD refers to the percent modified core recovery (Deere, et al.,


1967), calculated from drilling logs by deleting from the "recovered"
catagory all pieces of core less than four inches long*. A minimum

* Barton, Lien and Lunde state that Norwegian Geotechnica l Institute


geologists have found they can estimate RQD values in jointed, hard,
clay-free rocks from field estimates of the number of joints per
cubic meter ( J ) . RQD = 115 - 3.3 J (RQD < 100)
v v
Goodmao-Gt ol oglcal Eng - 5
52 Rock Classification

RQD of 10 is used in evaluating Q. The other terms in Equation 5


evaluate the number of joint sets and the roughness, alteration,
water and stress conditions according to Table 2-11.

TABLE 2-11

Values of the Parameters


in Barton, Lien, and Lunde's Classification

A. Number of sets of discontinuities In


massive 0.5
one set 2.0
two sets 4.0
three sets 9.0
four or more sets 15.0
crushed rock 20.0

B. Roughness of discontinuities Jr*


non-continuous joints 4.0
rough, wavy 3.0
smooth, wavy 2.0
rough, planar 1.5
smooth, planar 1.0
slick, planar 0.5
"filled" discontinuities 1.0
* add 1.0 if mean joint spacing exceeds 3 meters

C. Filling and wall rock alteration Ja

a) • essentially unfilled

healed 0.75
staining only; no alteration 1.0
silty or sandy coatings 3.0
clay coatings 4.0

b) filled

sand or crushed rock filling 4.0


stiff clay filling <5 mm thick 6.0
soft clay filling <5 mm thick 8.0
swelling clay filling <5 mm thick 12.0
stiff clay filling >5 mm thick 10.0
soft clay filling >5mm thick 15.0
swelling clay filling >5 mm thick 20.0
Rock Classification 53

Table 2-11 (continued )

D. Water c o nditions Jw

dry 1.0
medium water inflow 0.66
large inflow with unfilled joints 0.5
large inflow with filled joints
which wash out 0 .3 3
high transient inflow 0. 2 - 0 .1
high continuous inflow 0.1 - 0.0 5

E. Stress reduction class SRF*

loose rock with clay-filled


discontinuities 10 . 0
loose rock with open discontin u -
ities 5.0
rock at shallow depth «50m)
with clay-filled
discontinuities 2.5
rock with tight, unfilled dis-
continuities under medium
stress 1.0

* Barton et al also define SRF values corresponding to


degrees of bursting, squeezing, and swelling rock conditions.

TABLE 2-12
After Barton, Lien, and Lunde (1975)

Q Rock mass qual i ty for tunnelin g

<0.01 exception a lly poor


0.01 - 0.1 extremely poor
0.1 1.0 very poor
1.0 4.0 poor
4.0 10.0 fair
10.10 40.0 good
40.0 - 100.0 very good
100.0 - 400.0 extremely good
>400.0 exceptionally goo d
54 Rock Classification

Barton's analysis of case histories yielded a relationship for


the maximum safe span (D) for an unsupported underground opening as
a function of Q:

D 2.1 (Q)0.387 (6)

where D is in meters, and Q is in the range 0.001 ~ Q ~ 1,000.


Other functions of Q are given to select supports for different types
of openings.
For example, consider the rock masses in Figures 2-9c and 2-ge
with respect to tunneling at 40 meters depth. In the former case,
assume the RQD is found to equal 30% and in the latter 75%. Assuming
there will be no water inflow, we might estimate Q for each case as
follows. For the rock of figure 2-9c:

1.0 1.0
Q
2.0 0.83 (7)
2.0

According to Table 2-12, this classifies as very poor rock; the


maximum unsupported span according to (6) is about two meters. For
the rock of figure 2-ge:

75 1.5 1.0
Q 5 (8)
15 1.0 1.5

This qualifies as fair rock; the maximum unsupported span is about


four meters.
Barton's classification scheme has considerable potential for
engineering for underground works as well as for generalization of
experiences in other areas of engineering. A somewhat similar
classification, developed by Bieniawski (1974) is presented in Table
2-13. Of course, no classification system can assign a name as
generally informative as a careful description of the geological
environment, the rock material, the weathering profile, and the
system of discontinuities. Table 2-14 summarizes those factors
appropriate in a geotechnical description of a rock mass. Table 2-15
Rock Classification 55

is a standardized data sheet developed by the South African Central


Scientific and Industrial Research Organization (CSIRO) and is useful
for providing input for Bieniawski's classification.

TABLE 2-13
Geomechanics Classification of Jointed Rock Masses

A. CLASSIFICATION PARAMETERS AND THEIR RATINGS

POlnH oa d Use 01 unl d x lat c u rl'


Strength 8 MPa 4 - 8 MPa 2 - 4 MPa I 2 MPa
streng !l ) Ind\3x pre SS l v e leS T p r e l p r l(~ d
01
IntCict rOck
material
Uniaxi a l
comp r es sive
s tr eng th
200 MPa 100 - 200 MPa 50 100 MPa 2S - 50 MPa 10-b
MPa
l ..l-HI
M Pd
I ' J
M P"

15 12 2 j 1
1 0

Dr il l core quality ROD 9 0 ~', - l ()O%. 75". - 90 % 5 - 75 ", 25"" · 50'" Z'I ' I

20 17 13

Spacing 01 JOints · 3 m , -3 m 0.3 - 1 m SO - 300 mm ~ (I mm

30 25 20 10
~+-------------------4-------------~-------------+-------------+-CS;I~ic
~ -en-S'~
d-ea~s-u~rla-c-eS~S-0-1t-9-0l-
, g-
e --~-
n-'M-'~-(~
~

Condition 01 jO i n ts
ve~O~o~;~II~~r~~~es S~~~~::I~~,;h s~rr~,,:s S~;~~I:a;~oUngh s~r~~es : :Gouge :, m m IhlCk OR
No seperatlOn Hard JOint wall ro c"" Soft JOint wall rock JOints open 1-5 mm JOints open 5 mm
Hard JOint wall rock Contin u o us j Oints Continuous JOints

Rat in g 25 20

Inllow p el 10m 2:5 2S 125 1::<',


N one IJire s.'nl!n !It (es. fl i I
tunne l length
~--~~~-10R -------------------------' OR OR OR---- - -- -
JOl n l w ;Hf'1
Gr o(;r d I
waler RallO maZ~~~;~~CI~ol 0.0 U2 0.2 - 0 "
~----~~-10R-------------------------~ OR OR OR
MO ist o nly Wat e r uncer ~o oeraie Lle" r'l('
General conditions Complelely dry (Inl e t S Ifl al water) pr essu r e l/"d l ~1 I.."ll n l· ·nh

10

B ADJUSTMENT FOR JOINT ORIENTATIONS

Strik e and dip Very V ,\ ,


Favou.' a t)le- Fair Unfallourable
orl e n! at l/l ns o f JO int s lavourable un ilit 1. 1' 011 '

Tunnels -2 -5 -10 'i

Ratings FourlCJ attOns -2 -7 · I~

Slopes -5 -25 -50 -60

C ROCK MASS CLASSES AND THEIR RATINGS

Cla ss NO II I IV

D e-sc fl p t, o n Verv good rock Goo o ro c k

100,- 90 90 ..... 70 70 50 SO ,. 25

D. MEANING OF ROCK MASS CLASSES

Class No III IV

Average stand-Up time 10 year s ! o r 5 m s p an 6 m o n l ns l o r 4 m span , w ee k for J m 5pan

Cohesio n of (h e rock mas s 300 kPa 200 - l aO ,Pa 150 - 2Q!! , P a 100 . 150 , Pa 100 . "'"

C r l ~(I On a n g le 0 1 (he rock 'lla ss JO· .J5 .'0


W dl not c ave readi ly
C .a\jl eabi ll ty 01 ore V ery p o o r Fa ir
Larg e frag me nts
56 Rock Classification

TABLE 2-14
Some Factors to be Considered in
a Geotechnical Description of a Rock Mass

A. Rock material

Petrologic description -- rock name, texture, fabric, principal


and accessory minerals; nature of cement; alteration effects.
Presence of alterable minerals such as gypsum, pyrrhotite,
etc. should especially be noted.

Classification as "rock',' "weathered rock" or "soil-like rock"


according to results of simple tests (see Table 2-1).

Weatherability according to slake-durability or other test.

Mechanical properties according to an index test -- e.g.


Schmidt hammer, point load test, or scratch hardness.

Degree of weathering according to laboratory index tests or


mineralogic criteria.

State of fissuring, determined from polished sections or thin


sections or by results of wave velocity measurements,
tension tests, volumetric compression, or radial permeabil-
ity tests.

Micro structures in the hand specimen -- bedding, foliation,


etc.

B. Weathering Profiles

Description and classification of all the intermediate


weathering products and their spatial arrangement together
with results of laboratory tests indicative of their
mechanical properties.

Description of joint properties in the different stages of


weathering.

C. Discontinuities

Preferred orientations and spacings of each set, structural


name, (e.g. bedding, joint) for each set; roughness
angles versus wave length and description of wall rock
as wavy, rough, smooth, or slickensided; note roughness
anisotropy.

Wall rock scratch hardness expressed by a standard termin-


ology, or strength as measured by Schmidt hammer.
Rock Classification 57

Table 2·14 (continued)

Filling material: thickness; completeness of filling;


c ompactness; composition; % clay and soil properties;
classification as: swelling, erodible, soluble, or
or stable.

Interlocking and tightness of fit: healed, 'close, open,


cavernous (or loosened).

Other features: estimate of relative extent; chemistry


of wate r; wIl l rock alteration.

TABL E 2·15
Input Data Form:
Geomechanics Classification of Jointed Rock Masses

INPUT DATA FORM: GEOMECHANICS Cl ASSIFICATION OF JOINTED ROCK MASSES


Name of project
Site of survey
STRUCTURA L ROCK TYPE AND ORIGIN CONDITION Of JOINTS
Condu cte d b y REGION
CONTINUITY Sel I Sel2 Sel3 Set 4
Dale
I NOI continuous, no gouge
D H ll l CORE OUA LIT Y ROD W EA T H ERING
With gouge
Ve ry go o o Q!.. ah ly 90 100 Unweath ered Continuous rJQ gouge
G ood qu ality 75 - 90% Slightly w ea tne red
With gouge
Fair quailly 50 75... Moderately weathered
SEPARATION
Poor quality 25 50% Highly wealhered
Very tight tOlnts Less than a 1 mm
Ver I poor Q lI <:il lty 25 0;0 Compl e t ~ l y w e a ther ed
N O T e RO D Rock Ouailly D e slg nallo n Tight JOlnlS a t - 1 mm
STREN GTH O F IN T A C T ROCK MAT ER IA L
,n ,1cearda n"e Wl lh Ihe mel h od o f Dee re
Moderately open JOints 1 - 5 ln m
GR O U ND WA TER U r "d x< a l POin t ·t o act
_ _____ _ _ _ _ _ _ Deslgnallon c om p r e ~~ lve
slre n gth Open JOints More than 5 mm
Index
INF L O W per 10 In Illres,'rnlr""\ute st re ng U' ROUGHNESS
of tunn el longlh Very nigh Over 200 M Pa B MPa
Very rough surlaces
' hgh 100 - 200 MPa 4-8 MPa
WA rE R PRESSURE, kPa Rough surfaces
Medium 50 lOa MPa 2 - 4 I"IPa
25 50 MPa 1-2 MPa Sllghlly rough surlaces
GENERAL C O N D ITIO NS (completely dry. mOist Low
Very low 10 25 MPa 1 MPa Smooth surfaces
only water un der pre SS'He, se\lere problems)
3- 10 MPa SlickenSided surfaces
~________________________~__________~J_
M_p a~____________ -4 JOINT WALL ROCK

S P A CIN G OF J O IN T S Hard rock


Set I Sel 2 S~ t 3 Se l4 Medium hard roc~

Very wide Over 3 rn


Salt rock
Wide 1 J m NOTE Provid e data lor e ach 10 lnl ~ ~

Mod8rf'tlely close 03 - 1 m M AJOR FA ULT S OR f OLD S


50 - 300 mrT'

V c r 'I Clo se SO mlT'l


NOTF 1 h r s ,~ vah H~ S a re tl bta l n d from rt Jom t survey and no t tram borehole logs
F' rO", ld r- d d ta 101 eact) )o .n t se t _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _+ D
_ e_Sc l o_r l_
_ r_'b_e _m_a_ a Ll~ S n d fOld s specifYing the ir loc_a ~ ty , natu re and o r lentCt~ - -- -1
f-- - -
STRIKE AND DIP ORI E"' TA T IO NS GENE R AL R EMARKS AND ADDITI O NA L DATA
} - --
-~--;;-;;;;;' nt sPecli y -;;-;; lype,-ihl ~ ;;;;;~ S , contlnully and c o n s ls1enc y
Sell Strike (Irom 10 DIp
Describe waviness of JOints
(average) (angle) (dlr~Cllon) Assess regional stresses
Strlkt=:' IIrom 10 DIp

S l ri k e ( tr am Iu Dip

Stnk!:: (from 10 D'p


NOTE The' data rm 1I1lS form cunstltlile the nllllllllum rr.qlllr~d for (mqlne II " Q lI eS fll ll
Prov .rlP data fOf ft,l ,~ h tOlr l sel Re te' ::}_II _d_" e_c_II_on_S_I'_'1_"_il0_n_el_'c_o_o_ftr_ _ _ _ _L-_ _ _Tt---.:'e...:g:..:.eo~lo___.:g~IS~1,_-h_,'_UI---.:d,_r,_ow_e_ve_r_s---.:uP:..:.P___.:lv_a_nY:..:.I_lll_ll_le_,'_"l_ul_-m_a_llo_n_w_h_'C_h_tlL'"_C_OI_'i_'d_L'_
" _'8_lo_,V'_"1...
'
3
principles of
stereographic projection
and joint surveys

CONFORMAL STEREOGRAPHIC PROJECTION

Problems in geology and engineering which involve relationships


of lines and planes in space can be solved or simplified through
reference to stereographic projection. This is a method of mapping
the surface of a sphere onto a plane, used in crystallography,
cartography, navigation, structural geology, geophysics and other
fields. By using the stereographic projection alone, one can solve
problems involving orientations of lines and planes such as determin-
ing the plunge of the intersection of discontinuities. Where the
position as well as the orientation of a line or a plane is involved,
such as in analysis of overturning in a potential rock slide, the
stereographic projection must be supplemented by other constructions.
In this chapter we will consider the basic constructions necessary
for applications in geological engineering and an introduction to
statistics with reference to joint surveys. Applications in photo-
grammetry, the kinematics and statics of rock blocks, orientation of
structures in drill cores, and assessment of the state of stress in
rocks will be considered in later chapters. With these applications
in mind, this chapter provides a thorough consideration of the
properties and techniques of stereographic projection, including basic
constructions.
Figure 3-1 illustrates the geometric principles of stereo-
Stereographic Projection and Joint Surveys 59

graphic projection. Any point on the surface of a sphere is pro-


jected onto a diametral plane of the sphere (the projection plane) by
means of construction lines radiating from a focus; the focus is
fixed at one of the ends of the perpendicular to the projection
plane. In figure 3-1, the projection plane is horizontal and the
focus is at the bottom of the sphere. Points in the upper hemisphere
will therefore appear inside the projection of the horizontal plane,
which forms the Irprimitive circle lr
, a name derived from crystallog-
raphy. Line ON , for example, whic~ pierces the sphere at NR in the
R
upper hemisphere projects to point n on the horizontal diametral
R
plane. An inclined diametral plane, which intersects the sphere as
a "great circle", projects as a circular arc as shown inside the

Trace 0 f pl ane A

Upper hemisphere
pro j ec tion of A

Trace and projection


of horizontal plane
('primitive circle")

Figure 3-1. Stereographic projection of


points on a reference sphere.
60 Stereographic Projection and Joint Surveys

primitive circle. This arc represents the half plane in the upper
hemisphere; the completion of the circle in the projection plane
beyond the primitive would represent the projection of the half plane
in the lower hemisphere. This fact derives from the theorem, proved
in Phillips' text on Crystallography (1946) that the projection of
a circle on the reference sphere is a circle on the projection plane .
The projection is conformal; angles measured between lines on the
sphere are preserved by the projection. As will be seen later, a
variant of the projection exists that loses this quality (the "equal
area" projection).
In dealing with problems of orientation only, i.e. where the
spatial positions of lines and planes are not considered, one may
move a line or plane parallel to itself to pass through the center
of a single reference sphere. Any plane, then, can be described by
the projection of a great circle, and any line by the projection of
a piercing point on the reference sphere. A great circle, by conven-
tion, will be assumed to represent a plane, and a point will represent
a line.
A vertical section of the reference sphere, figure 3-2a, provides
a complete description of the geometric relationships basic to the
method of projection. An upward directed line OP through the center
of the reference sphere will appear in the upper hemisphere projec-
tion at a radial distance from the center equal to r tan a/2 in the
direction of its bearing, where r is the radius of the reference
sphere and a is the complement of the angle of rise. The opposite
(tail end) of the line (-OP) projects to point -p outside the primi-
tive circle, at a radial distance equal to r cotan a/2 in the opposite
direction. If -op represents the dip vector of a plane, points p and
-p are points on the great circle representing its stereographic pro-
jection. The bisector of line (-p) (p) is the center of this great
circle which may be constructed with a compass, figure 3-2b. In
this figure, the plan view of the projection plane and the vertical
section of the reference sphere in the direction of dip have been
superimposed. It will be found that the center, q, is the same point
as the stereographic projection of a line plunging in the direction
opposite to the dip at an angle from vertical equal to twice the dip
Stereographic Projection and Joint Surveys 61

~.
2 r ctn cr/2 .. I
r ctn a/ r tan cr/
2

-p t

focus
(a)

distance pr := distance qo

stereographic projection
of sma ll circle of 20°
about E

-p

stereographic projection of
inclined plane through p and N
(dipping 20°)

primitive circle

(b)

Figure 3-2. (a) Vertical section of the reference sphere through line PO. (b) Upper hemisphere stereo-
graphic projection (plan) superimposed on the vertical section of the reference sphere through I ine PO.
62 Stereographic Projection and Joint Surveys

angle. The center can also be located graphically, as shown in


figure 3-2b, by the intersection of the projection plane (OE) and a
perpendicular to the dip vector through the focus (line FQ). It
will also be useful to project small circles on the reference sphere
formed by the double cone of the lines making a fixed angle with a
given axis. For example, in figure 3-2a, we consider the cone formed
by lines making an angle of 90-a with the horizontal (OE); two of the
lines of this cone are OP, and ~T. The bisector of the projections
of these lines, points p and t, is point r (figure 3-2b) which is the
center for construction of the required circle. When the axis of the
small circle is horizontal, the center r can be found as the point
of intersection of the projection plane OE and a tangent to the
reference sphere at P.

Stereonets

Making use of the construction for the centers of great circles


and small circles about a horizontal axis given in figure 3-2b, it
is possible to construct a family of inclined planes having a common
horizontal intersection and a family of small circles about this
intersection, creating an "equatorial" stereonet. (Wulff net, figure
3-3a). If one regards the axis of the small circles (the intersec-
tion of the great circles) as the polar axis, (i . e. the focus of
projection is on the equator) then the family of great circles can
be viewed as the projection of the lines of longitude and the small
circles as the projections of the lines of latitude. The equatorial
net is convenient for the tracing of planes and the measurement of
angles from tracings, as shown in figure 3-6. When dealing only with
lines in space, it may be easier to use a net constructed about a
focus on the pole. The resulting "polar" stereonet shows the lines
of latitude as circles constructed about the center, while the lines
of longitude are straight lines, i.e. vertical planes, through the
center. By marking any point on the primitive circle as north, a
line may be plotted by the intersection of the required bearing line
and small circle of plunge. For example, in figure 3-3b, point A
is 20 degrees from horizontal to the North 20 E. (The angle is above
or below the horizontal according to the choice of hemisphere). It
Stereographic Projection and Joint Surveys 63

3 50 10

270 90

lA O

(a)

Figure 3-3. (a) Equatorial conformal stereonet; (computed by


Dr. C. St. John; reproduced with permission). (b) Polar conformal
stereonet; (computed by Dr. C. St. John; reproduced with permis-
sion).
64 Stereographic Projection and Joint Surveys

280 80

n. IIIIII~·
260 100

180

(b)
Stereographic Projection and Joint Surveys 65

is useful to print a supply of tracing sheets having a polar confor-


mal net lightly drawn upon them for use as overlays on the equatorial
net. Lines can then be read directly from the tracing. (Until
familiarity is gained with the basic constructions, however, it might
be better to use a clear tracing to reduce the number of lines and
ar c s) .
For statistical discussions of orientations, it is preferable
to use a modification of the stereographic projection which produces
eq u al areas for a given solid angle subtended anywhere on the sphere.
It will be noted, in figure 3-3a and 3-3b, that a "square" bounded
by two de g ree variation of latitude and longitude grows smaller
towards the center of the projection. An equal area projection can
b e developed from the stereographic projection as shown in figure 3-4.
I magi ne a f o cus a t the bottom of the reference s p here and the pr~­

jection plane tangent to the reference sphere at the top. The


stereograp hi c projection referred to this plane becomes larger than
the one referred to the diametral plane, but remains geometrically
similar to i t . The equal area projection of line OP, is obtained by
swinging an arc from N (the top of the sphere) to point P, giving

1---- -- /2r - - - -- I

Figure 3-4. Equal area (Lambert) versus equal angle (Wulff) projections. The distance
from the center of the reference sphere to the projection of OP (referred to plane OE)
V2
is r tan (cx!2) in stereographic projection and r sin ((d2) in equal area projection.
66 Stereographic Projection and Joint Surveys

point P2' So that the radius of the resulting primitive circle will
be r, it is necessary to divide the projected lengths by 1:2 giving a
distance from the center of 1:2 r sin (a/2) for a line plunging at 90-a
from horizontal. Equatorial and polar equal area nets (Lambert nets)
are given in figure 3-5a and b. The equal area projection is not con-
formal and projections can not be made using circular constructions
as for the stereographic projection. However, all the basic con-
structions required can be performed on a tracing by sketching lines
from the nets given, exactly as presented for the stereographic
projections in examples to be discussed.

BASIC CONSTRUCTIONS

In a number of examples basic manipulations with stereograms will


be demonstrated, first by construction, and then by use of a tracing
on the conformal equatorial stereonet, as depicted in figure 3-6.

Figure 3-6. Use of the stereonet with a tracing.


Stereographic Projection and Joint Surveys 67

180

(a)

Figure 3-5. (a) Equatorial equal area stereonet; (computed by


Dr. C. St. John; reproduced with permission). (b) Polar equal
area stereonet; (computed by Dr. C. St. John; reproduced with
perm ission).
Goodman-Geological Eng --6
68 Stereographic Projection and Joint Surveys

350 10

BlilliEF

280 80

270 90

260 100

190 170
180

(b)
Stereographic Projection and Joint Surveys 69

Construction of a line in the Upper or lower Hemisphere

Figure 3-7a reviews the steps for projection of a line. Given


a line rising 20 degrees to the N30 E, draw a circle of convenient
radius, mark north and locate the bearing on the primitive circle,
point B, by measuring 30 degrees from north, subtended at the center
O. For purposes of construction, draw a line through OB and a
perpendicular to OB intersecting the primitive at F. We may view F
as the focus for an upper hemisphere projection (an upper hemisphere
projection was chosen because the required line is in the upper
hemisphere). OB is the trace of the projection plane and the required
line OP can be drawn 20 degrees above OB. Line FP intersects the
projection plane at p, the required point. The opposite to OP can
be plotted similarly by continuing OP in the opposite sense to point
-P on the primitive circle, whence line (F)(-P) yields the required
point -p as it intersects the projection plane. Points near the focus
will project an inconvenient distance beyond the primitive and the
focus itself can not be plotted. However, this is really no limita-

N Plot line p rising 20° to N 30 E

plot opposite -p

Figure 3-7. (a) Stereographic projection


of a line and its opposite.
70 Stereographic Projection and Joint Surveys

revolve tracing to place point B


on equator; measure angle
by counting great circles

Figure 3-7. (b) The


problem of figure 3-7a
SOUTH using the stereo net.

tion because one can switch the hemisphere, or work with the opposites
of inconveniently located pOints without difficulty. Figure 3-8
shows the relationship between upper and lower hemisphere plots of
a point and its opposite. In many problems of interest, as for exam-
ple in joint surveys, structural geology calculations, and core
orientation calculations, we do not distinguish a point from its
opposite and all required constructions can be confined to one hemi-
sphere. However, in statics, where we associate a direction with
certain lines, the entire sphere must be mapped. In this case, as
will be shown in chapter 6, we may use a single continuous projection
with focus selected carefully to reduce the size of the figure, or
we may use two side by side projections, one for the upper hemisphere,
Stereographic Projection and Joint Surveys 71

FocuS for LH projection

Figure 3-8. The relationship between


upper and lower hemisphere plots. FocuS for UH projection

marked U.H., and the other for the lower hemisphere, marked L.H.
Figure 3-7b shows how the tracing can be used to facilitate the
plotting of a line. Mount the stereonet on cardboard and push a
thumb tack (drawing pin) through the center so that the tracing may
be revolved about the center (protect the tracing with a piece of
transparent tape to prevent enlargement of the hole with use). Mark
point B 30 degrees from the arbitrarily selected north position and
revolve the tracing to superimpose B on the equator. The angle of
rise, 20 degrees, is measured along the vertical plane represented by
the equator, counting ten great circles. The opposite to point p
can not be plotted using the net (the opposite to the lower hemisphere
projection of p is shown).

Plotting a Plane and its Normal

Figure 3-9a shows how to plot a plane given its strike and dip.*
Plane R strikes N 20 Wand dips 13 degrees to the NE. The strike
vector, SR' is horizontal so lies 20 degrees west of north along the
primitive circle. The other end of the strike vector (-SR) is on
the opposite end of the diameter through SR. The dip vector D and
It

* The terms strike and dip are discussed in chapter 4, page xxx
72 Stereographic Projection and Joint Surveys

plane R

primitive

plot plane R and its normal nR in


the upper hemisphere:
strike: N 20W
dip: 13° to NE

Figure 3-9. (a) Stereographic projection of


a plane from its strike and dip.

(a) NORTH
SOUTH U.H. 20,0__--"'---_

nR 0

cente~r O~30
for great ~
circle
26 0

(b) The problem of figure 3-9a using the


stereonet. LH (b)
Stereographic Projection and Join t Surveys 73

its opp os ite - DR are constr uc te d as i n fi gu r e 3-7a. To construct t he


plane, simply draw a circle throu gh SR ' - SR' DR ' and - DR ' It wil l
be helpful to remember t h at t he ce n ter fo r t hi s ci rcle is at t he
intersection of : t he trace of the p roj e ct i on p lane (ODR ); and the
perpendicular to the l i ne of d ip (OP ) through -SR as shown.
The normal to a p lane R i s fre que ntl y req ui re d . This may b e
done by findi n g the d ip vector DR in t he pl an e , re covering the l ine
of dip in vertical s e c t ion throug h OD , a nd pro je cting t he p ierci n g
R
point NR of the perpe n d icular to i t e r ected a t O. F or example , i n
figure 3-7a, if we h a d been g ive n p oin t p i n th e st e re ographic pro -
jection, we could h ave const r u c ted poi n t F, and tr aced l ine Fp to
determine point P , in th is wa y re c over i n g t he actual li n e OP which
point p represents. A simpl er way to p lo t t he no r ma l to plane R wit h
strike and dip vecto r s project i ng at po in t s SR an d DR ' figu r e 3-9a ,
is as the intersection of diamete r OD R an d a l in e making an angl e of
45 degrees with DRS R or -DR-S R as s hown .
To plot a plane and its normal us i n g the s te reone t and t racing ,
mark the strik e an d dip vecto rs as i n f i g ure 3-7b, and t race the
great circle passing through both; s in ce SR is h o rizontal , this great
circle will be foun d by revolv in g the tr a c ing till SR is o ve r th e
polar position. To improve ac cura cy in prob l e ms inv olving c ons t ruc-
tion of great circles, the cir cl e shoul d be c ons tr ucted wi t h a compass
rather than tr aced fre e han d . Th e cent er is in s i de t he net f or planes
dipping less than 45 de grees; as sh own , the c en t er i s l ocated as a
point 26 degrees from vertical in t he d irec tio n opposed to t he dip.
(The angle from verti ca l to t h e ce n ter is a l way s t wice t he amo unt
of dip.)

The Angle Between Two Lines

The angle b et wee n two l i n e s ( r e membe r i ng t h e conve ntion that al l


lines pass thro ug h t he ce nte r o f t he r e ferenc e sphere) is meas u r e d in
their common plane. Give n two li n es p an d q , ( f i g ure 3-l0a) , f ind
the opposite t~ eac h and con s tr uc t t he g reat circ l e p q ( passing
through al l f o ur poin t s) . Co n struc t t h e normal n
pq
t o plane pq a s
in t h e previous examp le, figure 3- 9 a. Now d r a w st ra igh t li nes npq p
and npq q, in te rs ecti n g th e pr imi t i ve c i rcl e a t pOints P1 and q1
74 Stereographic Projection and Joint Surveys

Find plane common to two lines NORTH


p and q
plane pq
Measure the angle between
p and q

p plunges 24° to N26W


q plunges 26° to N83W

Figure 3-10_ (a) Stereograph ic projection


to measure the angle between two lines.

NORTH
(a)

,plane M __--..::.....

(b) Solution to the problem of


LH figure 3-10a using the stereonet.
Stereographic Projection and Joint Surveys 75

respectively. The required angle from p to q, is the angle P10ql'


subtended at the center of the primitive circle.
This problem is easier on the stereonet, figure 3-10b. Plot
points p and q and revolve the tracing until they lie along a common
great circle. Angle pOq is measured along this great circle, which
is calibrated by its intersections with small circles.

The Line of I ntersection of Two Planes

Two planes will project as two circles, which intersect as two


points, figure 3-lla. These are the intersection lines in each
hemisphere. On the stereonet, figure 3-llb, only one of the inter-
sections will be given, (the other being its opposite ) .

NORTH

primitive

Find line of intersection


in lower I'lemisphere, '
between planes Rand Q
R strikes North, dips 45W
Q strikes North 30E,
dips 15°NW
Figure 3-1 L (a) Stereographic projection to find
the line of intersection of two planes.
76 Stereographic Projection and Joint Surveys

Figure 3-11. (b) So:ution to prob-


LH lem of 3-11 a using the stereo net.

Orthographic Projection of a Line on a Plane

This construction is used to determine angles in perspective


views, and in other applications as illustrated in later chapters.
Figure 3-12a shows the method of solution, given line p and plane R.
Construct the normal to R (n ). The orthographic projection of p in
R
R (line J) is the intersection of the great circles nRP and R. The
attitude of J is most often referred to plane R; in this case J
pitches* 16 degrees below horizontal from the north in plane R. The
solution on tile stereonet is given in figure 3-12b.

* An angle of pitch, or rake, is an angle measured in a non-vertical


plane.
Stereographic Projection and Joint Surveys 77

(a) Find orthographic projection


of p on plane R:
p plunges 17° to N 31 E
R strikes N37E, dips 43° NW
J is the required line

Primitive
Figure 3-12. (a) Finding the orthograph-
ic projection of a line on a plane.

NORTH

c
R-

(b) Solution to the problem of


3-12a using the stereonet. a (b)
78 Stereographic Projection and Joint Surveys

Construction of a Small Circle and Its Opposite

We use small circles in two ways. Problems involving rotation


about an axis involve movement of points along small circles con-
structed about the axis. Problems with friction involve construction
of the cone of static friction about the normal to a surface of
sliding. The projection of a small circle is shown in figure 3-l3a,
where we desire a small circle of radius 15 degrees about the axis p.
Draw a diameter through p and its perpendicular through 0 to locate
the upper hemisphere focus F, as in previous examples. The line Fp
intersects the primitive at P so that line OP is the axis of the
double cone which forms the small circle on the reference sphere.
0
Lines 15 above and below OP yield points PI and P , which project to
2
PI and P2 along Ope The required small circle is drawn with center
at the bisector of PI and P2; (the center coincides with the axis
only if it is vertical). The opposite of the small circle about p is
constructed similarly as a circle through the opposites, -PI and -P2'

Find small circle


of "rad~us" <p
about line p.
NORTH
p

p rises 33°E
cp = 15 0

primitive
circle

UH

Figure 3-13. (a) Construction of a small circle and its opposite.


Stereographic Projection and Joint Surveys 79

NORTH
~l

(b) Solution to the problem of 3-13a


using the stereonet. UH

A small circle and its opposite may range over any portion of the
plane.
The construction of a small circle on the stereonet tracing is
illustrated in figure 3-13b. Points on the required small circle are
obtained by revolving the net to pass a sequence of great circles
through p, counting the required number of degrees in each direction
from p in each great circle in turn. Of course, if the conformal
stereonet is used, one can superimpose the construction of 3-13a on
the tracing to improve the accuracy of the drawing. If the small
circle lies partly in each hemisphere, as in figure 3-l3c, it may be
satisfactory for some applications to show two arcs, one being the
80 Stereographic Proiection and Joint Surveys

ap,b + (- b)(-P2)( -0)

describes the complete


set of non-directed Ii nes
200 from p.

Figure 3-13. (c) A small circle in both


F hemispheres and its opposite.

opposite of the incomplete portion of the other (this is not satis-


factory for problems in statics).

Rotation of a Line About an Axis

A small circle defines the locus of rotation of a line about an


axis. Figure 3-14a shows an example in which a line p is rotated 30
degrees about a horizontal axis q. The path of rotation is found by
constructing a small circle about q as in previous examples. This
small circle can be calibrated as follows. Construct the plane R
perpendicular to q. Locate point I in this plane by its intersection
with the great circle common to p and q. Angles of rotation about
q are measured in plane R. The required point p is found by
r
locating m 30 degrees from I in plane R, and constructing the inter-
section of great circle mq and the small circle. The construction
is identical when the axis of rotation is inclined, as in figure 3-
15a.
On the steronet, one can immediately trace the path of rotation
using the family of small circles, as shown in figure 3-14b, when
the axis of rotation is horizontal. When the axis is inclined, the
construction of figure 3-15a can be performed on the tracing, or the
problem can temporarily be rotated so that the rotation axis becomes
horizontal. The latter technique is illustrated in figure 3-15b.
Stereographic Projection and Joint Surveys 81

(a) R, the plane -L to q


NORTH

RotatG p count e rc lockw ise


30° about a
horizonta l axis q
to positi on Pr

ce nter for sma ll circle

p rises 2?D to N33E


q is East

Pr is 51 0 to N49E

small ci rcle about q through p,


the locus of p in rofatio n
about ct
U.H.

(b) NORTH note: rotati on f~om . p to Pr ~s s ~ ow n


IS clo ckWise If projec tion IS L. H.

Figure 3-14. (a) Rotation about


a horizontal axis. (b) So lu ti on to
the prob lem of 3-14a using the
stereonet.
82 Stereographic Projection and Joint Surveys

(a)
small circle about <l
through p, the locus of
p in rotation about q

Rotate p counterclockwise
120° about q to position Pr

p rises 45° to North


q rises 45° to East

Pr rises 37° to S7E

NORTH (b)

o
U.H.

I
45 0 --- q
0
45 ---

Figure 3-15. (a) Rotation about


an inclined axis. (b) Solution of
the problem of figure 3 -15a by
an alternate method using the
stereonet.
Stereographic Projection and Joint Surveys 83

First q is rotated to horizontal (q' ) along the equator carrying p


to p , along the small circle through p. Then the tracing is
a .
revolved, to place q' over the polar position and the required angle
of rotation about q is traced along the small circle through p ,
a
carrying it to point Pb' Finally, Pb is carried to its correct
position p as q' is returned to q.
r
Problems involving rotation of data arise in solution of geo-
logical structures, and in analysis of symmetry of fabric data. An
example will be discussed later (figure 3-16) in connection with
joint surveys.

JOINT SURVEYS AND STATIST ICS ON TH E SPH ER E

Field measurements of joint attitudes will define a scatter of


orientations wh ich can be interpreted statistically if plotted on
an equal area steronet. Turner and Weiss (1963) discuss various
means of contouring the resul ting "scatter diagram" on the equal
area diagram to highlight the symmetry of the discontinuity system.
The "preferred orientations" of joint sets are defined by "highs" of
the contour diagram. If many points are to be plotted and contoured,
it might be appropr iate to use a computer program with an automatic
plotter routine, s uch as given by Jeran and Mashey (1970). Plotting
and contouring can be done by hand using a polar net for plotting and
a counting circle with radius equal to l/lOth of the net radius. The
number of points inside the counting circle, when placed arbitrarily
on a scatter diagram, is written in the cen t er of the circle. By
ranging over the hemisphere, pole density values are defined to a
detail necessary to sketch contours. If two counting circles are
provided exactly one net diameter apart, one circle will be the
opposite of the other and points on the edge of the net can be
counted by adding t he number in the two circles.
Statistical distributions on the sphere have been discussed by
Arnold (1941), and Fisher (1953) an d applied to contoured joint
diagrams by Pincus (1953), McMahon (1972), Mahtab et al (1972) and
others . A problem basic to applications in geologi cal engi n eering
is finding the p robability of occur r ence of a joint normal in a
given orientation, once the preferred orientations have been deter-
Good m J -~ G f olog ' cal Eng. -7
84 Stereographlc Projection and Joint Surveys

mined. For this type of work, the precision of the counting circle
technique of contouring is questionable and one may consider using a
vector approach as discussed by Mahtab et al (1972). The mean orien-
tation of a group of unit vectors (joint normals) clustered within
a cone is determined by the orientation of the resultant R of all
the unit vectors in the cone. If all the joint normals are exactly
parallel, the magnitude of R
will equal the total number of normals
in the cluster N; as the joint directions become dispersed, the
length of R becomes less than N. Thus, as an index of dispersion of
orientations, one can use a coefficient K defined as follows*

N
K (1)

K approaches infinity as the dispersion of joint normals approaches


zero. This definition of K fits Arnold's and Fisher's hemispherical
normal distributions which, for K greater than 6, yield the proba-
bility formula:

cos 1JJ 1 + (11K) In (1 - P) (2 )

P is the probability that a normal to a joint occurs within 1JJ degrees


of the mean orientation. It will be called the "probability of
1
occurrence". The standard deviation of the distribution is 0 = (K)-2
A rigorous statistical analysis must include an examination of
the degree to which the hemispherical normal distribution fits the
data; the application of the chi square test for goodness of fit is
discussed by Irving (1964), and Mahtab et al (1972). A simple
approach can be based upon the comparison of dispersion coefficients,
K, fitting different probability contour levels. Suppose we are
given a cluster of normals; first, estimate the orientation of the
mean attitude as the high point of the set of contours drawn using

* The absolute value of it is used to permit replacing a pole by


its opposite in the calculation.
Stereographic Projection and Joint Surveys 85

a counting circle on a n equal area plot. * On a separate trac ing,


figure 3-16 sketch a curve about the me an e nc los i ng a l l but 5 percent
of the number of points in the cl uster ; t h i s i s a n esti mate of the
ninety-five percent probability curve ( P = 0 . 95) . Then s k etch a

NORTH

IJtso =9. 5 to 12°

1j195 = 18.5 to 22°

for IJlso =11 °, K = 39

Figure 3-16. Probabil ity on the sphere. fo r '/195=20 .5 ° , K=47

* The example of figure 3- 16 is on a confo rma l n et, but an equal


area net might have been use d.
86 Stereographic Projection and Joint Surveys

closed curve about the mean orientation containing one half of the
total number of points, estimating its position in such a way that
the excluded points are equally distributed around the curve, if
possible; this is one estimate of the fifty percent probability curve
(P = 0.50). Rotate the diagram to orient the mean vertically (figure
3-16) . If the shape of the curves after rotation is very different
from a circle, an anisotropic Fisher distribution can be used, i.e.
one in which K varies with direction (see Shanley and Mahtab, 1975).
If the shape approximates a circle, as in figure 3-16, measure the
angle ~ from the mean orientation to the 50 percent and 95 percent
probability curves and compute K corresponding to each value from
figure 3-17, derived from the expression

K
In (1 - P) (3 )
cos (~) - 1

If the values of K.50 and K.95 are not greatly different, the
hemispherical normal distribution can be used to predict angular
distances from the mean corresponding to other probability values
(the positions of the 50 and 95 percent probability curves as drawn
can be used directly whether or not the distribution fits the hemi-
spherical normal distribution).
0 0
In the example of figure 3-16, ~50 is in the range 9.5 to 12
0 0
while ~95 is in the range 18.5 to 220. For ~50 = 11 and
0
~95 = 20.5 , figure 3-17 gives values of K respectively equal to 39
and 47. If we retain the value K =
39, we could predict for example
l 0
that the 99% probability angle (~99) is cos- (1 + In (0.01/39» = 28 .
Given the mean joint orientation in a design problem involving poten-
tial sliding, one might then consider the "design joint" to be
o
inclined 28 steeper than the mean attitude. The suggested approach
will be, however, to retain the mean joint orientation in calcula-
tions and reduce the angle of friction accordingly, as discussed in
Chapter 6.

BIAS IN MEASUREMENT OF JOINT ORIENTATIONS


FROM DRI LL HOLES AND OUTCROPS

Ruth Terzaghi (1965) demonstrated that joint spacings inferred


from measurements of joint attitudes in boreholes or from planar
Stereographic Projection and Joint Surveys 87

30.-----~----~~------r_----,_----~--------._----_r----~------~

25
0

~
20
C-
o
(l)

E
(l)
..c
15
E
~
'+-- 10
~
(J'l
c:
«
5

0
5 10 20 50 100 200 500 1000 2000 5000
Dispersion Index, K

Figure 3-17. Angles of 50% and 95% probability poles from the mean of a hem ispherical normal distrib ution
After Arnold (1941), Fisher (1953), and Mahtab et al (1972) .

outcrops are biased because joints pa ral lel to the surfa ce or the
axis of the borehole are not seen. Let a be the an gle between the
normal to a joint and the axis of a borehole or t h e n orm al to a
surface outcrop, (figure 3-18). The act ua l number of intersections
with the outcrop or borehole by a se r ies of pa ra llel joints wi ll
depend on a. Only when a equals 90° will the true spacing equal the
outcrop length divided by the number of joints; similarly, onl y when a
equals 0° wil l the spacing equal the drill hole length divided by the
number of joints. However, the data can be cor re cted as shown i n fig u re
3-18. Let N be the number of joints obse rved at an gle a. We can the n
a
calculate a "correct" number of joints Nc as follows.
For borehole measurements:

N
c
Na Icos a (4a)

and for measurements from a planar outcrop:

N Nisin
a
a Ia I (4b)
c
88 Stereographic Projection and Joint Surveys

a.
s

True spacing between jOint planes:


Outcrop - S = Lo sin a
Dnllhole - S = LH cos a

Corrected number of joint planes in length L :


Nc = ~; Outcrop - Nc = _ _L_ = Na . _._1-
S Loslna. slna

Drillhole - Nc= __L__ = Na:·-I-


LHCOS a cos a

A Na in counting
~. Circle

(':1a~ Drill Hole


a§- ~
Normal to
Outcrop

~n
Surface
counting "Blind zone"
circle -20 0 < a < 20°
Figure 3-18. Correction
for bias in joint surveys;
after Ruth Terzaghi (1965).

The indicated limits on application of these corrections suggest that


the measured data may be too sparse in the "blind zones" of the bore-
hole or outcrop (figure 3-18). Normally, joint observations are
made from compound surfaces and there is no absolutely blind zone.
However, there still may be bias. The spacing of sheeting joints
in granitic rocks and massive sandstones, for example, are difficult
to appreciate from outcrops.

THE DIRECTIONALITY OF A JOINTED ROCK MASS

Talobre (1957) suggested a simple construction to determine


S tereographic Projection and Joint Surveys 89

f undamental directi on s of a join ted rock mas s. If the preferred


orie nt atio n s of j o i n t sets a nd ot he r d i s c on tinu ity s ur faces ar e
plotte d as po les on the ste r eone t, on e can t ry to fi n d a position
whi ch makes an ang le wi t h e ac h p o le of less t han the f ric t io n ang le
¢ (figure 3- 19). Such a direction wil l be i n the area common to small
circles of radius ¢ a b o u t each pole ( ea c h joi nt set may present a
d iff ere nt an g l e ¢ ).
I f an a re a common t o all the sma ll circles e x i s ts, it wi l l c on -
t a i n the s tro ngest d i r e ctions of the rock mass because a r ock lo aded
in such an orient a ti o n can mobil i z e rock st r e ng t h witho u t j o i n t sl i p .

12
13 14

Figure 3-19. Constructi on


fo r t he d irecti o n d~) of
greatest compressi ve strengt h
in a jo inted rock mass; after
Talobre (196 7) , chapter 10.
90 Stereographic Projection and Joint Surveys

Let b signify such a direction. A drill hole along b will intersect


A

a great number of joints, and b will accordingly be a good direction


in which to drill drain or grouting holes. If a tunnel is oriented
A

in direction b, the walls will be in the least deformable directions


of the rock mass, while the face of the tunnel will be in the most
desirable orientation for mining. See also Lokin (1974).
explo ation of
rock cond-t-ons

Most construction in rock entails adapting to existing terrain


a n d rock conditions. With adequate description of the morphology
and properties of the different rock classes, the attitudes and
characteristics of discontinuities, the hydrologic and sometimes
tectonic regimes, it is possible to choose feasible sites and to find
workable designs for engineering developme n ts in discontinuous rocks.
This chapter concerns the engineering geology methods used to obtain
the required knowledge -- geological mapping, air photos and geo-
physics, and exploratory excavations and drill holes. Attention will
be directed to methods pertinent to characterizing and mapping dis-
continuities. The review of principles presented here can be
supplemented by reading in Simpson (1968), Donn and Shimer (1958),
Badgeley (1959), Blyth (1965), Ragan (1974), or other works in
structural geology and map interpretation.

GEOLOGICAL MAPS AND THEI R IN TERPR ETATION

Existing Maps

Exploration should always begin with a thorough review of exist-


ing geological maps and reports. It must be appreciated, of course,
that all such maps are to a greater or lesser ext ent inte r pretive,
the degree depending upon the scale and purpose of the map. Maps at
scales larger than 1:500 can be almost entirely observational and
Figure 4-1. (a) An engineering geology map derived from a 1: 10,500 general purpose geologic map. From
the Geol. Soc. of London, Engineering Group Working Part Report, 1973; (courtesy Prof. W. Dearman).
(b) A portion of the legend for figure 4-1a.

grade into logs, but these rarely will be found. Available general
purpose geological maps show the mapper's opinion, based upon incom-
plete evidence, on the distribution of mappable bedrock formations
and structures below the mantle of soil. The surface distributions
of mappable formations are shown by means of the surface traces of
their contacts with adjacent units. By means of structural symbols
and supplementary cross sections, subsurface information is presented
as well. Such maps are basic for all rock engineering.
Exploration of Rock Conditions 93

(b) SO LID FORMAT IONS


CARBONIFEROUS
Upper Carboniferous (Namurian)

~~:\:~elr~~~~~y,F~~~~~~~~,
G kF th1n 1y bedd ed to th i n ly
lami n ated, very cl osely jo i n ted, slig h tl y to mo derately
we at hered, poo rly c leaved SHALE, wea k , imp e rmeable excep t along
open joints . Inter bedded wi th ve r y s ubo rdina t e grey to dark
greenish grey, fine-grained , ve r y t h i n ly bedded, th i nly lami nat e d
and cross-laminated, clo sely joi nted, s l ig ht l y to moder at el y
weathered SILTSTONE, moder a t ely s tro ng, a nd dar k gr e enish grey
medium grained, very thinl y to med ium be dded , wi th c lo se ly to
widely spaced joints slig ht ly to modera tely we athere d , SAN DSTONE,
strong.
The shale slakes on expo su re and is sui tab le for brick making.

~ ~t~~~~~~~~ le
predominates.
to map gr ou ps of beds in whi ch SA NDSTONE
Beds a r e usua lly less t ha n 12 in . thic k an d are
separated by very thin beds of s iltstone and s ha l e .
Sand stones are suitable for a ggr e ga te production.
Within the contact metamorphic aureole of th e gr a nit e , dar k gr ey ,
very pale orange to dus ky yel lowish br ow n , f i ne t o me di um g r ain ed,
thinly bedded, close ly jo i nted, s light ly t o moderate ly weathered,
hornfelsed SHALE and SAND STONE, strong , i mpe r vious e xcep t along
open joints. Local l y with fine gr ain e d bl a ck t o urmal ine
developed as selvedges up to 1 in. wi de a l ong di s co nti nuities and
with irregular quartz veins up to 2 i n. wide.

Geological maps can be obtained f rom fe deral, st at e an d local


governments, geological societies, theses a nd journals. The U . S ~ G . S .

offers indexes by states. State geological s ur veys a l so supply their


own indexes to maps. Also, the Commission for t he Geo l og i ca l Map of
the World publishes periodic reports on wo r ldwi de ma ppin g in t he
Geological Newsletter, (Internation a l Union of Geological Societies).
Such maps can be helpful for engineer ing plann ing. Sometimes more
detailed engineering geological maps exist a nd can be ob taine d fr o m
government or private engineering o rgan i zati ons.

Geological Mapping

Detailed geological mapping f r om surface expos ures , reinf orced


by trenching, and other subsurface e xp loration, is t h e b a sic veh icle
for developing the model of rock con d itions. The mappi ng s ca le
should be selected to delineate necess a ry details. Sho ul d t he re su lt-
ing map size be too large, suppleme n t ary notes can be in co rporat e d,
as for example in figure 4 -1 from the Geolo gi cal So ciety of London' s
Engineering Group Working Party Report (1 972 )*. The map wa s pre pared

* Quarterly Journal of Engineerin g Ge olo gy, Vo l 5, no . 4.


94 Exploration of Rock Conditions

by annotation and expansion of detail from a previously existing


published geological map at a scale of 1:10,560; potentially ambiguous
names were replaced by objective descriptions and annotations using
standard terminology and symbols defined in the Working Party report.
Such a procedure minimizes the risk of misinterpretation, which can
prove costly in engineering construction projects. Figure 4-2 is
a portion of a bedrock geological map prepared by an engineering
geologist (originally at 1:2400 scale with 10 foot contours) for
design of a large earth dam. The accurate location of contacts,
faults, landslides and structural attitudes in relation to project
features required special mapping reinforced by coring and trenching
through the soil.
Site maps and logs for engineering design should show the limits

400

~~5CXJ

?
550----'''

\
2== != N

o 100 200ft.

Scole

Figure 4-2. Geologic map of a portion of the left abutment of an earth dam;
(courtesy Kaiser Engineers, Oakland).
Exploration of Rock Conditions 95

of outcrops and the locations of borings and trenche s to separ ate ,


wherever possible, the interpretative from the observational data.
However, even directly observed features are interpretive ow ing to
the nature of the geologic mapping process whi ch entails f ixin g not
only the location but the classification of the str at ig rap hy and
structure. A geological map is prepared by charting the correct
geographic location for sets of points on known geologic horizons
and structures. For example , given a succession of l ay ers consist ing
of (A) sandstone, (B) limestone, and (e ) shale, the limestone is
mapped by plotting the trace of the contacts AB and BC. If a fourth
layer (D) consisting of limestone underlies the shale, a decision will
have to be made at every outcrop of the shale/limest on e contact to
insure that Be and not CD has been located. Usually a contact such
as BC can be located only approximately as a line (dashed) between
proximate outcrops of B and of C. The quality and usefulness of a
geological map depend markedly on the previous e x perience of t h e
maker, who needs to be accurate s imultaneously in his geog rap hi c and
geologic "locations".
Special surveying is rarely necessa ry for general geological
mapping but for site mapping required accu ra cy may demand survey of
some suitably located control points. Rarel y will a ccu r a c ies in
location of contacts be demanded better than one foot in plan o r
elevation, as geological uncertainties and variations will usually be
more limiting.
The basic f i eld tool is a combination c ompass and c linometer
which is used to measure the attitudes of rock planes, to resect field
positions by horizontal sightings, and to estimate elevations from
vertical triangulation. The elar compass*, allows si mu ltaneo us
reading of strike and dip directions of planes in contact wit h the
instrument. It is less versatile and more expensive than the better
known Brunton compass** (or "pocket transit") which permits h o ri zontal
and vertical sightings as well as attitude measurements on planes in

* Breithaupt Co., Kessel, W. Germany


** "Brunton compass" of Ainsworth and the "pocket transit" of K&E are
two examples.
96 Exploration of Rock Conditions

contact with the case. By installing a protractor along the side to


indicate the rotation of the cover it will be possible to read
strikes and dips simultaneously. Golder & Brawner & Associates,
Vancouver, B.C., and the U.S. Bureau of Mines (Bolstad and Mahtab,
1974), both have developed non-magnetic dip and strike instruments
with azimuth measurements referenced by sightings on a control point.
This is necessary in some iron ore mines and quite frequently in
underground construction projects.
As additional data become available through exploratory excava-
tions and drill holes, the geological map is improved. During con-
struction its details can be checked as the subsurface is exposed
to view. An opportunity to gain experience in understanding geologi-
cal mapping is afforded by the collection of instructive British
maps prepared by Blyth (1965).

Mapping Planar Features

Outcrop patterns of planar discontinuities and non-folded strata


reflect the geometry of intersection of a plane and an uneven surface.
The orientation of a plane is determined by the bearing and plunge of
any two mutually perpendicular lines it contains; one of these lines
--the strike vector -- is usually taken along the horizontal direc-
tion and the other -- the dip vector -- is then pointed down the
plane (figure 4-3). The bearing of a line is the azimuth of its
orthographic projection in a horizontal plane; the plunge of a line
is the vertical angle between the line and its projection in the

Di p vector
of plane p

Figure 4-3. The convention


for strike and dip.
Exploration of Rock Conditions 97

horizontal p l a n e. It is suffi ci e nt t o state: the b eari ng of th e


strike vect o r (the "strike") , and the plun ge of the dip vector (t he
"dip") toget her with an indication of the quad r a n t in which its
0
bearing lies, for example: st ri ke N30E, dip 20 SE. Table 4- 1
presents strike and dip symb o ls f or va ri ous planar f eatures; when
located on t h e map, these s ymb ols fi x a plane in o r ientation and i n
position. Some prefer to repres e nt the p l ane b y the beari n g and
p l unge of the d ip vecto r; others repor t the bearing and plunge o f a
plane's normal (the "pole " ). In both cases, the user must un derst and
that the given line represe n ts a p l ane.
Representati on of a d iscontinuity surface b y a plane al lows
projection of dat a away from points of observatio n . Fo ur types of
projections ar ise: ( 1 ) t he strike and d ip of a pla n ar feat ure a r e
known a t a po in t on the surface and t h e trace o f the feature across
the uneven groun d surface is to be determined; ( 2) t he surfa ce trace
of the plane is k nown a n d the strike an d dip are to b e dete r mined ;
(3) the strike is known b ut not the dip, a nd the sur f ace t ra ce is
incompletely known, for example on l y at two outc rop s; a nd ( 4 ) t he

TAB LE 4-1
Symbols fo r Geological Maps (courtesy U .. Geol. Survey )

40
• Contact, sh ow ing dip ...... ,~
Norma l fault; hachures on downthrown si de
90 10
I Contact, vertical (left) and overtu rned ....L. Strike and di p of bedding
Con tact, located ap prox . (give limits) ~ Strike and dip of overturned beddin g
Contact, located very approx. 90
--+ Strike of vertical bedding
,
90 Fa ult, showing dips (!) Horizontal bedding
10 20
Fa ult, located appro x . (give limits) -A. .Jo,., Und ulatory or c rumpled beds
60
- --?- -- Fa ult, existe nce uncertai n I St ri ke and dip of bedding, uncertain
Fault, projected beneath m ap ped units 70
...L ~? St rike of bedding certa in but di ps uncer ta in
Possibl e fau lt (as located from aerial Stri ke and dip of foliations
photographs) Stri ke of vertical foliati ons

<! 75 0
u
Fa ult, showin g tren d and plunge of
linear features (0, down-thrown
+++ Hori zontal foliations
Stri ke and d ip w here bedd ing
side; U, upthrown side) para llels foli ati on
Fau lt, show ing relative hor izontal Strike and dip o f jo ints (left)
move ment and veins or d ikes
Thr ust faults; T or sawteeth in Strike of vertical jo in ts (left)
u p pe r pl ate and veins or dikes
Pault zon es, sh owi ng ave . d ips Horizontal jo ints (left) and vei ns or dikes
98 Exploration of Rock Conditions

Figure 4-4. Projection of data from an outcrop.

amount of dip is known but not the strike, and the surface trace is
known only at several points. Figure 4-4 shows a strike and dip
symbol plotted at the outcrop of a clay seam. Based upon the measured
plane, denoted stratum 1, one can estimate the trace of the clay seam
across the ground surface, which can then be checked by excavations
through the soil. Since a line parallel to the strike of a plane is
a contour on the plane, the clay seam should also underlie the sur~

face at A, B, C and D where the strike line intersects the 20 meter


surface contour. A vertical section in the direction of the dip is
shown on the right of the map. The line A, B, C, D intersects this
section at elevation 20 (pt E) as shown. The edge of stratum 1 can
be drawn in this section because the dip angle, 45 0 , has been given.
Line A, B, C, D, E may be termed "the 20 meter structure contour".
Structure contours at other elevations can be drawn from the cross-
section, and their intersection with the appropriate surface contours
Exploration of Rock Conditions 99

define additional p o ints along t h e t ra ce o f t he clay s e am acros s t he


surface.
Suppose geol og i c a l map pi ng had re veale d suf fi c ie n t a ctua l ou t-
crops to de fine th e t r a c e j u s t c on st r ucted yet wi t hout yi e lding
anywhere a precise measuremen t of st rike and d ip . I f t h e t race pas ses
more than once t h r o ug h t h e s a me con t our , the s t r i k e is de termi n e d by
the trace. Li n e DA i s par a l le l t o t he s t r i ke. Interp re t i ng l i n e
DA as a structure conto u r at 20 met e r s an d l i ne FG as a st ruct ure
contour at 10 mete r s e l evat i o n , t h e p lan e c a n be con stru c t e d i n
vertical se ct i o n by exte n d i n g l i n e EH d e te rmin ing t he dip.
Occas ionally, some p oint s a l o ng t he s u rf a ce t race a r e known b u t
too few to a ll ow the con st r uct i on abo v e . If the st rike i s known , t h e
dip can be de termi ne d , all ow i ng c omp le t ion o f t h e outcrop t r ace . For
example, i n f ig u r e 4- 4, if o n ly A a nd G were iden t i fi ed as po i nts o n
the trace of st r a tum 1, k nowl edge o f t h e st r i k e would allow con -
struct i on of points E an d H and t h e b a l ance o f t h e so l u ti o n f o l l o ws .
Following explo r ati on by d r il l i ng, one usu a lly k n own t he amount of
dip (the compleme n t o f the core/ b e ddi ng a ngle i n v ertica l drill ho l es)
but not neces sa rily t h e strike. In this case o ne mu st s e l e c t on e
of two possible solu t ion s, a s sh own i n f i g ure 4-5 . Here s t ratum 1
0
outcrops at G and B. The k nown d i p of 45 g ives a 1 0 me t e r hori z ont a l
offset distance (c o r r es pondi ng to t he 10 me t er di f fer e nce in e l e v a -
tion) which is laid o f f as a c i rcle about B. The s t ri ke li ne i s a
tangent to t h is circle drawn t hrough G. It is un iq uely defin e d as
line GC if we k now the di re cti o n o f s t rike i s approximate ly EW.
If 3 poin ts are l o c at ed o n t h e s t r a t um at diffe r en t e lev at ion s ,
the strike and di p are ob ta i n able . Poi n t s G, B , and K on t he t race
of stratum 1 (figure 4- 5 ) p resent such a 3 point p r obl e m. In ter -
polate between the h i gh est a nd l owes t pain t s (K and G) to find the
position whe re the st ratum wi l l h ave an el evat ion correspond i ng to
the middle poi nt (B ) Th is y iel ds poin t X (t he s trat um i s 10 m abo v e
the g r ound a t X, me an ing it h as been e roded off). Li ne BX , th e
20 meter structure conto ur , d e f i nes t he dir e c t ion o f st ri ke. The
dip is de termin e d by t he pro j ecti on of G o r K into the cross- se c ti on
as before or f rom the pe rpen dicula r distance d f rom G t o EX, ( s ince
the elevation di ffe r e nce f rom l in e BX t o l ine CG is known t o b e 10
Goodman- Geo logical Eng - 8
100 Exploration of Rock Conditions

Figure 4-5. Construction of a cross section


from outcrop data.

meters) .
In these examples one observes how the trace of an inclined
plane forms a "V" pointing in the direction of dip upon crossing a
valley. The trace of a horizontal plane will follow the contours
while a vertical plane will pursue a straight course in any terrain.
A plane inclined mo~e gently than the relief provides a more complex
and often multiple trace across the surface.
Cross-sections to convey the geological structure can be drawn
in any direction by projecting surface points parallel to the strike
into the desired section. In figure 4-5, a vertical section has been
Exploration of Rock Conditions 101

drawn through Sl and S2' The apparent ang le of dip in the section
diminishes as the angle (a) between the section and the strike
approaches O. Section Sl S2' in figure 4-5 with a = 42 gives an
apparent dip (AD) of 33.8 0 . It could also have been calculated from
the relationship derived from figure 4-6.

tan (AD) = tan (D) sin (a) (1)

where a is the angle between the strike and the section.


The spacing (t) between parallel planes is defined in the direc-
tion of their normals. This direction is contained in only one
vertical section, i.e. the one perpendicular to the strike. The
direction perpendicular to the edge of planar features in any other
vertical section will give an apparent spacing t', derived from
figure 4-6 as follows:

t' (2)

In figure 4-4, a second planar surface 20 meters beneath stratum 1


has been drawn on the true dip section and the surface trace of
stratum 2 has then been constructed as shown. The area between the
two traces maps the area underlain by the stratigraphic interval
between 1 and 2. In figure 4-5 the trace of the lower surface is
seen to intersect the line of the oblique section at L. The apparent
spacing, t', found by constructing an apparent dip throug h L, is
23.5 m, which agrees with equation 2, (a = 42 0 and D = 45 0 ).

Application of these simple principles will enable you to predict


where excavations and foundations are likely to encounter important
planes of weakness a n d to layout site plans so as to minimize the
difficulties such discontinuities can invite.

a structure contour

Figure 4-6. Apparent spacing (t')


and apparent dip (AD) in vertical
sections not parallel to the direc-
tion of dip. h st ructure contour
102 Exploration of Rock Conditions

AERIAL PHOTO INTERPRETATION

The opportunity to observe, and even measure, discontinuities


in the three dimensional model generated by viewing overlapping
vertical aerial photographs with a stereoscope can provide informa-
tion on the relative orientations, spacings, and extents of dis-
continuity surfaces. Figure 4-7 presents several examples to
illustrate how such material can be obtained. (These figures should
be viewed with a lens stereoscope.) Figure 4-7a shows tilted sedi-
mentary strata -- Mesozoic sandstones and siliceous shales -- in
the Colombian Andes. The zig-zag pattern reflects resistant beds
and the rule of "V's" discussed in the previous section can be used
to determine the direction of dip. A set of transverse joints (T)
crosses the beds without displacing them. Figure 4-7b shows sedi-
mentary strata in a desert environment in North Africa. Again a
set of recurrent fractures crosses the bedding, but in contrast to
figure 4-7a, these do offset the strata, showing that they are minor
faults. In both examples, the spacing of the fractures can be estab-
lished if the scale is known. The scale ratio, e.g. 1:20,000, for
any point equals the focal length of the lens (f) divided by the
height of the camera above the point (H). In figure 4-7b, the
average distance between shears is of the order of 0.01 feet and the
photograph was taken with a 6 inch lens from a height of 30,000 feet,
giving a spacing of 600 feet. A detailed study will indicate a dis-
tribution of spacings.
Figure 4-7c contains a dip slope defended by sandstone from
which a bed of shale has been almost completely stripped away. The
beds are widely spaced but two sets of major, apparently open
fractures spaced at 100 to 200 feet (the scale is 1:20,000) disrupt
the continuity of the formation. The fracture orientations are
regular but there is a gradual swing in the preferred orientation of
one of the joint sets from one side of the photo to the other. At
(1) in figure 4-7c a joint surface is exposed on which a strike and
dip measurement may be obtained. Presumably the joints visible in
the landscape are only a small fraction of the total population,
which must include members spaced several feet apart. However, the
(d )
(e)

(c)

Figu re 4-7. (a) Ve rt ical aerial photo (stereo-pai r )


of sedimentary strata in the Colomb ian Ande s.
(b) Stereo-pair, showing m inor fau lts; North Afr ica.
(c) Stereo-pair, showin g join ted sa ndstone, Wyoming.
(d) Stereo-pair, showing fract ures in sandsto ne,
Wyoming. (e) Stereo-pair, showing f ractures in
granite, South East Alask a.
104 Exploration of Rock Conditions

100 foot spacing figure may represent a reasonable estimate for the
spacing of ~ey discontinuities. Field study would tell. Observe
that the lengths of particular individual fractures are more than
1400 feet. Figure 4-7d presents a similar example but one in which
spacings are more variable. Again, two approximately orthogonal
sets are expressed in the terrain. The Northeast trending joints are
deeply etched in a closely spaced zone. Figure 4-7e presents a
fracture system representative of granitic terrains. There are at
least three different sets with considerable scatter about the mean
orientation of each, as well as considerable variation in length.
Study of aerial photographs may be the only reasonable way in many
instances, to gain data about joint lengths.

Measurements of Attitudes of Planar Features in Aerial Photos

Strikes and dips of discontinuities like those shown in figure


4-7 can be estimated from their outcrop trace and surface expression,
but one must be careful about the exaggeration of relief, usually
between two and three times, when making such estimates. From the
discussion of geological mapping it will be evident that any dis-
continuity lineament which presents a straight trace across uneven
topography such as (2) in figure 4-7e must be dipping very steeply.
On the other hand, surfaces such as (1) in figure 7e may be inclined
because their outcrop trace ariross the t~rrain is not straight.
The geometry of the vertical aerial photograph is such that
points on a horizontal plane are in correct map position with respect
to one another. The scale of the map changes with elevation but
horizontal lines are not rotated. This makes it easy to measure
strikes of line elements in the photograph with respect to a reference
direction. The usual reference -- the x axis -- is the line connect-
ing the principal point (the lens center) and the image of the princi-
pal point of the adjacent photograph (the conjugate principal point).
(The best way to view two overlapping photographs stereoscopically is
with their x axes aligned). Figure 4-8 shows the image of two points
Rand Q in adjacent, overlapping vertical photos. To determine the
bearing of the line RQ relative to the x axis, simply read the angle
with a protractor. Determination of the plunge of the line between
Exploration of Rock Condit ions 105

Fiduci al mark

Yo
YR
----,0
_ ---, R I TR
TO
I
I I I I
I I I I

0
. t- x +x'

XR = 1.15 X/R= - 2.50 o IS princ i pal poin t of left photo


xQ = 2.35 X/O= -0.95 0' IS principal point of right photo
YR = I 40 PR = 3.65 o and 1
0 are locate d on left an d r ight
Yo = 1.85 pO= 3.30 ph oto respectiv el y by Intersection of
lin es th rough fiduc ial marks
dPRO = 0.35

Figure 4-8. Parallax and coordinates of points in ove rlap ping vertica l aerial photographs.

points Rand Q requires that we first ca lcu late t he i r paral l ax


difference, dPRQ

(3 )

where, as shown in figure 4-8, the primes de n o te p o i nt s i n t h e ri ght


photograph an d th~ unp ri med coordin a tes r efe r to p o ints i n the l e ft
photograph. Th e e l e vation di ffer en ce b etwe e n R and Q i s (Moffi t t ,
1959 )

dh ( 4)
RQ

where HQ is th e fly in g height a b ove p o i n t Q. An approximat ion,


106 Exploration of Rock Conditions

usually suf~iciently accurate for strike and dip measurement is:

(5)

where H is the camera height above the principal point, and b, the
photo base, is the distance between the principal point and the con-
jugate principal point (00'). Both Hand b are averaged from the
two photographs. Let SRQ be the distance from R to Q measured on the
photograph. The ground distance between Rand Q is approximately
equal to SRQ(HQ/f). Thus, the plunge (S) of RQ is given by

(f) (dP)
Tan S RQ
(6)
(SRQ) (PQ)

Using the average photo base b and flying height H, if the point Q
is at about the same elevation as the principal point,

Tan S ( 7)

Since the determination of parallax difference, dp, involves the


difference in measurements of larger quantities, the measurements can
not usually be performed with sufficient precision using an engineer's
scale. The precision of standard stereometers, such as the parallax
bar, Coutour Finder, and Stereo Comparograph, is of the order of
0.01 mm. These instruments utilize the floating dot principle, in
which the fused image of dots over the right and left photograph
appears to rise or fall as the separation is changed while viewed
under a stereoscope. By fixing the dot "on the ground" at each point
along the opposite ends of a dip vector, the difference in parallax
. can be read from the difference in absolute readings of the instru-
ment. It is easier for some people to use a floating line, illustra-
ted in figure 4-9, with a "parallax wedge height finder"*, consisting
of two non-parallel lines with a separation in the range that includes

* Felsenthal Instruments, Chicago


Exploration of Rock Conditions 107

(a)

(b)

Figure 4-9. Measurement of dip using a parallax wedge: (a) th e floa t ing line pierces the ground at t he
bottom of t he sl ope; (b) the floati ng line p ierces the ground at th e top of th e dip slope.
108 Exploration of Rock Conditions

the average person 1 s eye-to-eye spacing. When placed over the stereo
pair and viewed with a lens stereoscope, the lines fuse to give the
appearance of one sloping line in space. In figure 4-9a, the sloping
line appears to pierce the ground at point Q at the top of a dip
slope. The separation of the lines on the wedge is 0.0130 feet.
In figure 4-9a, the wedge has been moved to make the floating line
pierce the base of the slope at R; the separation is now 0.0210
feet (estimating the last digit) giving a parallax difference dPRQ
of 0.0088 feet. The photo distance SRQ is 0.043 feet. The photos
were taken through a lens having a six inch focal length and the
photo base, b, averaged from the two photos is 0.317 feet (this can
not be verified on the stereopair of figure 4-9 as not enough of the
image area has been copied). Solving equation 7, the bed is dipping
at an angle of 17.5 degrees.
If the angle between the lines on the parallax wedge can be
varied while the model is viewed, the inclination of the line in
the model will appear to change. This is the principle of the
"parallax ladder"* for measurement of slopes. "Hackettrs method"
employs an analogue device to measure directly the exaggerated slope
angle in the stereo-model; the exaggeration factor is then obtained
by a graphical procedure as discussed by Ray (1960).
The most accurate measurements of attitudes will be obtained
using a plotting instrument such as a Kelsh plotter or Multiplex, in
wh~ch a tilt free stereo-model is created by projecting through film
or glass plate diapositives of the aerial photographs. The attitude
of a planar surface can be obtained by mathematical plane-fitting
to coordinates of three or more points on the surface. Ross-Brown
(1972) in fact even determined joint roughness by computing the
scatter of planes fitting four or more points in a stereo-model of
terrestrial photographs. Attitudes of discontinuity surfaces in the
model can also be measured by tilting the plotting platen into
conformity with the surface to be measured and reading the attitude
of the table with a clinometer just as it is done on the outcrop
(Ray, 1960).

* Photogrammetry, Inc., Silver Spring, Maryland


Exploration of Rock Conditions 109

While it is difficult to estimate dips reliably owing to vertica l


exaggeration , the eye can be trained by first measuring several
reference slopes in the model. For fracture analyses, it may be
satisfactory merely to classify each of the joint sets by strike
and approximate dip, e.g., nearly vertical, steeply , moderately, or
gently dipping, or nearly horizontal. The principles of geological
mapping reviewed in the previous section will enable such estimates
to be made from the outcrop trace. In fact, non-parallel lines in
a planar surface can be combined with an apparent dip solution, using
the stereograph ic projection (Wallace, 1950).

Types and Sources of Aerial Photos

Black and white vertical aerial photos like those in figure 4-7
have been taken of nearly every part of the United States by agencies
of the U.S. governmen t and are at t ai nable a t modest c ost. Table 4 -2
lists the principal sources of photographic coverage in the United
States. An order may be expedited by visiting the offices of the
agency and identifying the photo exposure numbers on an index mosaic.
Alternatively, the area desired in photos can be marked on a topo-
graphic map and forwarded with the order. Faster response can be
obtained through private aerial photo companies. Local government
agencies , such as the County Tax Assessor's Office, may possess
recent aerial photos, flown for them on contract, which might be
borrowed. For analysis of fracture patterns, high altitude photo-
graphs, e.g. 30,000 or 20,000 feet, are often more desirable than
the lower altitude photos, e.g. 5000 feet, used to make contour maps
for engineering design.
In military and space research, there has been considerable
interest in expanding the scope of photo interpretation outside of
the visible spectrum through the use of special films and artifici a l
imagery techniques (Cassines, 1972). Infra-red photographs bring
out contrast of wet and dry terrain. Color photography, though rather
expensive, can point out zones of alt e ration. Thermal imagery
produces a visual record from natural heat radiating sources, such
that warmer areas appear lighter in tone; thermal imagery can some-
times show up seepage along underground paths, subaqueous discharge,
110 Exploration of Rock Conditions

TABLE 4-2
Sources of Existing Government Aerial Photos

United States Department of Agriculture, Agricultural Stabilization


and Conservation Service, Aerial Photography Division, 45 French
Broad Avenue, Asheville, N.C., 28802 (for eastern United States) or
2505 Parley's Way, Salt Lake City, Utah, 84109 (for western United
States).

United States Department of Agriculture, Forest Service, Division


of Engineering, Washington, D.C., 20250. (This is the address for
general information. Orders are processed through regional offices.)

United States Department of Agriculture Soil Conservation Service,


Cartographic Div., Federal Center Building, Hyattsville, Md., 20783.

United States Department of the Interior, Geological Survey, Map


Information Office, Washington, D.C., 20242. (This is the address
for general information. Orders are processed through regional
offices.)

United States Department of Commerce, National Ocean Survey, Washing-


ton Science Center, Rockville, Md., 20852. (Formerly the Coast and
Geodetic Survey)

National Archives and Record Services, Cartographic Branch,


Washington, D.C., 20250. (Old government photographs are obtainable
from this source.)

United States Department of the Interior, Geological Survey, EROS


Data Center, 10th and Dakota Avenues, Sioux Falls, S.D., 57104. (This
office maintains information on ERTS satellite photographs and
imagery. )

National Air Photo Library, Surveys and Mapping Buildings, Room 130,
615 Booth Street, Ottawa 4, Ontario, CANADA. (This is the centralized
source for Canadian air photos)

Note: The military authorities in most countries have extensive


photo coverage which engineering organizations with a "need
to know" can somet imes borrow. (Figure 4-7a is aU. S. Air
Force photograph.)

and geothermal anomalies. Radar imagery (SLAR) is perhaps the most


valuable remote sensing technique for study of fracture patterns
(Barr~ 1969). In this technique, reflections from a radar beam
directed obliquely from an airplane are displayed on a cathod~ ray
tube, and photographed. Radar (and thermal) images will be obtained
Exploration of Rock Conditions 111

even at night and t h r ough clou ds . The l ow ang le Qf inc i dence,


compared to conventional high sun a n g l e photogr a phy a c c e n t uat e s
linear features such as faults. Cl uf f a nd Slemmons ( 1971) demonstra -
ted that conventional p h o tographs take n in the e arl y or l at e hours,
when the sun angle is low, will highligh t l in e a r f eatu res (fi g ure
4-10). In any of t h e s e dis p lays, t h e res o l ut i o n o f fra ct ure l in e a-
ments can be enhanced b y viewi ng the p ho t o through a f ilm wit h f in e
parallel rulings, e.g. 200 l ine s per in ch } he l d at a distance from
the photo; for when the ruled lines a re p e r pe ndicular to a li n e a r
feature, it can be seen to the exclus ion of li n es of a l l o t h er
orientations. Inf ormation on sources of r a da r imagery pi cture s and
other remote sensing techni q u e s can b e obtai n e d by writing to th e
EROS data center, U. S. Geolog i c al Surve y, (see Tabl e 4-2). Hig h
altitude and s a tel l ite p hotographs t ake n as part of t he American
space program can a l so b e consu l te d t hro ugh the Earth Reso urce s
Program of NASA, Lyndon Johnson Space Cen t e r , Ho us ton , Tex as.

(a) (b)

Figu re 4 -10. Low sun angle aerial p hotography will d isplay shadows al ong pro min ent discontinuiti es, such
as the Wasatch Fault scarp photographed here. (a) A convent ion al vertical aerial ph otograph of a p orti on of
the Wasatch front, Utah, photo grap hed at mid-day. (b) A port ion of t he same area, photographed early in
the morning. Reproduced from Cluff and Slemmons (1 971).
112 Exploration of Rock Conditions

TERRESTRIAL PHOTOGRAPHS

Photographs can be used to measure the orientations and spacings


of discontinuities in inaccessible outcrops and in transient exposures
such as cleaned rock foundations, mines, and exploratory trenches.

Measurements on Photographs

Using a 150 mm phototheodelite and good survey control, co-


ordinates of points 100 meters from the camera are determinable with
a precision of the order of 2 centimeters (Ross Brown and Atkinson,
1972). This is adequate for mapping open pit mines and will usually
permit sufficiently accurate determination of joint orientations
at inaccessible poin ts by "3 point" determination. (In practice,
Ross-Brown, et al. (1973) use 4 points.) Many engineers do not
realize that accurate measurements can also be made from photographs
taken with a hand-held camera. In connection with geological mapping
onto a suitable available topographic base and logging of explora-
tory excavations, ordinary snapshots can supplement quantitative
observations made in the field with a precision consistent with the
geological mapping process. An inspiring book on techniques appro-
priate for analysis of photographs taken with a hand held camera was
written by J.C.C. Williams (1969).
A photograph records an infinite family of convergent sightings
from the camera point to all points within the field of view. A
line from the lens center (0) perpendicular to the film plane estab-
lishes the focal distance (f) (figure 4-11), and intersects the photo
at the principal point (p) found on an untrimmed photo by the inter-
section of diagonals connecting corners. The nadir point (n) lies
beneath the lens center on the extension of the positive plane in
the case of a photo below the horizontal. (The zenith point (Z) lies
above the lens center on the extension of the positive plane on an
elevated photograph.) The obliquity of the photo is the angle Onp
(or Ozp) while its complement is termed the angle of rise or
depression (8). The horizon is a horizontal line in the positive
plane at the elevation of the camera; a perpendicular from the horizon
through p will pass through n (or z) and defines the trace of the
principal vertical.
Exploration of Rock Condi tions 113

'~--.--:::------I k
o = c amera stat ion
p = pri nci pa l point
kq' = hori zan
n ': nodi r point
L = isoce nter
Hori zo nta l ang le to q ;: H q :: k Lq '
Vertical a ngle to q :: Vq
Plun ge of camera ax i s ;: -1 k Op =8
Di stance s :
O p = foco l le ngth
p k = f to n 8
Figure 4·11 . Geometry of a Ok ,: k L ': flees 8
photograph. p n = f/ tonB

All points on the ground along the s a me ray to t h e le n s a re


superimposed on the positive . Dist a n ce s c a n be f oun d b y f ixing po ints
through resection from two pho t os . F i gure 4 - 1 2a, adapted f rom Mof fit
(1959), defines the co - o rdinates of a po in t o n a p hoto. The x-axis
is a line paralle l to the hor izo n th r ough t he p r inci p a l p o in t , p;
the principal vertical is the y axi s . ( A came r a de p re s s i o n an g le
is defined to be neg a t ive , gi ving a n egat ive va lue y = f tan s fo r
the principal point, p.) Let q be a p o i n t on th e posi tive. Pa i n t
q' is the projection of q on t he hori z ontal p l an e t hroug h the camera .
The horizontal angle, H , is kOq' given by
q

tan H f se c8 + y sin8
( 8)
q

the vertical angl e, Vq , is qOq ' given by


114 Exploration of Rock Conditions

q' (b) ,
(a)
\ \ L-------' f-'
J-\ I
\ ~ q,\
~ ----- I\
Hr
., HQ

\ ----
If::::
-- -c-.............

------ --- p- ~
i\...Horizon
n' \

\
~ I----- c-. . . . . . .~
~
\ v"-- ----
~\

~ r ~
\
L 1 -------- \ \

Figure 4-12. (a) Measurement of


horizontal angles on a photograph
by construction; points i and n are
described in figure 4-11. (b) Meas-
urement of horizontal angles on a
n tracing overlying a gnomonic net.

Yq cosB sinH q
tan V (9)
q xq

The angle Hq may also be found graphically as shown in figure


4-l2a. Moffitt gives a simple construction for vertical angles as
well.
A quick method of measuring angles on photos is through the use
of a gnomonic net. A photograph is a gnomonic projection wherein the
principal distance of the lens (f) equals the radius of the reference
sphere at p. Recognizing this, Wallace (1950) proposed superimposing
a suitably enlarged gnomonic net overlay centered on the principal
point of a photo to measure angles graphically.* Figure 4-l3b

* Wallace also showed how to solve for the attitude of a plane,


using the stereographic projection, from the angular relations of
lines in the plane as seen from two different perspectives. This
method is often inaccurate in practice, however.
Exploration of Rock Conditions 115

(a)

Posi tive
plane

o~----------~--------~~~

Figure 4-13. (a) A photograph as a gnomonic


projection. (b) A one degree gno mon ic net
corresp onding to a focal le ngth of 84.5 mm .

(b)
\ \ \ ,\\\\\\\\1\\11 \ TT1111111 1111 1 / 1 / / 1 / 1 I I
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ I \ \ I I \ \ \ I I I I I
\ \ \ \ \ \ \11\\ \\11 \JIILI I I I I
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \1111111 I I I I I
\
II \ \ \ 1111 I / / I
\
\ \
'\ \ \ \ TTl I 777

I
I
I
I
TTTTT!\\ \ I
°
I \
1 -r! / / I / / / , ff ! 1/ / I / I I / TlT!\\\\ 1 \ 1
lll T n l\
~H-I I I / / 11\ \1 1\ 11\ \\ 11\ \ \ \ \ \ \ \ \ \ \ \ \ \
II I II II I //II /II!I" I I I /I " / \1 \11111 1\\ \ 1 1 \ \ \ \ \ \ \ \ \ \ \ \ \ 1
I I / I I I -I 7 -I I / / / , I , , , , , I I I I I / / I I IT T1\ lT! \ 1 1 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
Goodman- Geologi ca l Eng-9
116 Exploration of Rock Conditions

Gnomonic projection
of a great circle - -
a plane through the
camera
y = f tan 8

Gnomonic projection
of a small circle of
latitude a
x = f ton 8
cos a
Figure 4-13. (c) Explanation of the gnomonic net.

0
is a 1 gnomonic net which can be superimposed on a photo if enlarged
or reduced accordingly. This net, plotted by computer, corresponds
to a focal length of 84.5 mm. (The focal length after any enlargement
will be found. by the relationship f y/tane where y is the distance
along the equator corresponding to an angle e.) As shown in figure
4-13c, the straight lines are the gnomonic projections of lines of
longitude -- great circles of the reference sphere having a common
intersection; they can represent any plane passing through the point
O. The curved lines are the projections of lines of latitude, i.e.,
small circles of the reference sphere; they graduate the straight
lines. To find the angle subtended at the camera between any two
Exp loration of Rock Conditions 117

imag e p o i n ts, r evo lv e the gnomonic overlay abou t p unt il o n e o f i ts


strai gh t l in e s (plane s) pas ses thro ugh bo t h po i n t s and coun t the
n umber o f l a ti tude c u rve s, e a c h of whic h ma rks an a n gl e o f 1° . An gle s
are alway s mea su red al ong the st ra i ght l in es by coun t ing t he c u rved
l in es. To me a sur e the hor i zon t al angl e Hq to an image po i n t, q ,
on ce t h e t r ue h or i zo n has been d r awn on th e phot ogr aph, revo l ve the
overl ay abo u t the pr i n c ip a l p oin t p to o r i e n t the s t r a igh t l i ne s
parallel to t he h ori z on a n d coun t th e a ng ul a r d i stances Hq as shown
o
in fi gure 4- 12b. Th e n rot a te the ove rlay 90 and me asure V. q
The
h or izon may be est ab lis hed usi ng t he gnomoni c net if the hori z on t al
an d vert i c a l ang l e s of t wo p oi nts in the photo f i e ld a re known, e . g .,
poin t s r a nd q i n f igure 4-14.
Th e above cons tr uc t ions r e quire you t o know the foca l length
(p ri n ci pal di stance ) of t h e c ame r a . Th e foc a l le ngth may be de-
t e r mined p reci s el y b y p hotog raphing a li ne of known l e ng t h a n d
positi on with resp ect t o t h e camera . An As ahi Pentax Spot ma t i c
wi t h a n omi na l 50mm l e n s foc u se d a t infini t y g a v e (f) = 50.5mm . A
wi de an g le l en s , nomi n a l ly 35mm, g ave (f) = 34 .4rnm .

A pplications of Terrestrial Photogrammetry

In r ec o nnaissance ge olo gical mapping, wh e re t he r e is rarel y


e nough t i me to measur e al l t he obs e rvat ions f rom e ach v i st a , an

r--.. ""'"- .....-


V
V ~
~ ~ ~

/ ~ -----" HO Izon vqt \


\~
I j"
q'

~
r- p
Vr
r--
~?

~ - ro-
I
Figure 4-14. Fi nd ing the horizon from known f-f-- I V
v-
vert ical angles t o two poin ts r and q . Rotate the V
r ~
t raci ng on t he gnomo nic net, about poi nt p, to ..-- ---... ..........
~
fin d an orientation giving correct values of vr and ~
vq . Then check the hori zontal angles Hr and Hq .
(see f igure 4~12b ) .
118 Exp /orCltion of Rock Conditions

oriented s nap shot i s v aluab le. Measureme n t s ma y be desired later


when an idea of t he project layout is proceeding a nd these can be
obtained without the necessity of revisiting every site. If a
topographic base map is a vailable, o utcrop limits, geological con-
tacts, springs, f a ults, p r ominent discon tin uities, a nd other features
seen in t he hills ides from a vantage point can be mapped onto the
base using the hor izont al and vertical angles f rom one camera station,
or the hori zo ntal a ngles from two camera station s. Photos from
helicopters an d light planes can also be used for this purpose if a
base map is avai l able and if promin ent points can be recognized i n
the photos. In this c a s e, t he camera point can be found by a t r ial
and error res ect i on us in g a tracing. If one is fortuitous in being
able to recognize a s u ff icien t number of images c orresponding to
points at kn own ma p position s, a method of alignmen t s, discussed b y
Williams, c an be use d to re se c t th e plan pos it ion o f the camera; any
two points on the same line pe rpendicular to the horizon establish
a ray to t he came ra an d two s uch rays intersect at the camera point.
With the pl a n p ositi o n of the camera point k n own, the vertical angles
to images recogn ized on the top ographic map will est a blish the
elevation of t he camera and the orientation of the principal point
can be estab li she d , etc. Thus, one can transfer geological details
from random ph otos to a topographic map.
In mapping f or desi gn stage studies, terrestrial photogrammetry
can be use d in two ways: loggi ng of exposures; an d mapping geological
details f rom in a c c essible pla ce s such as cliffs, rocks and walls.
Explorat ory e xca vat ions should be logged in comp l ete detail; a pro-
cedure makin g use o f s yste mat i cally taken photos will improve the
quality of th e loggin g. So ma n y s ty les of explo ratory excavation
present themsel ve s in practice that t he geologist will have to
establish the photo logging system for his own pu rp ose. Deep trenches
are difficu l t to photograp h effectively as a great many exposures
are requi r ed to c o ver the entire surface. Nevertheless, pho t o-
graphs are rel atively inexpensive, and the opportunity they afford
for later st udy an d measuremen t supports their routine application.
It should be possible to a rrange for multiple parallel photography
of the walls (each wall alternatively) of a trench or .adit, by
Exploration of Rock Conditions 119

mo unting th~ camera on a t ro l ley running paral le l to t he axis o f the


excavat ion on a wooden track; one can trip the shutter using a long
cable release.
Dur i ng construction and operation, . terrestria l photogrammetry
can be helpful in r e cording geological de tails during the short ti me
when new e xposures become accessible. Measurements can be made l at e r
a s neede d if the location and orientation of th e camera , and the
fo c a l l en g th of the lens , hav e been recorded.
Standard aerial photogrammetric techniques have been adapted t o
terres t rial photo t heodel i te photos for meas urements as a basis of
p a yme nt t o c ontrac t o rs* and for mapp i ng j o in t ori e ntat i on s in mi ne s
an d q u arries (Ross-Brown , 1972 and 19 7 3; Re ngers, 1967; and Linkwi tz )
1963) .
Sav age (1965) used aerial photo plotters with enlarged over-
l app i ng s t ereo photographs taken with an ordinary snap shot camera
from a le ve ll e d plan e table. Stereo pho tography is a helpful practice
eve n i n t hose cases wh e r e the parallax measureme nts wil l not form
t he basi s f or the photogrammetry because inspection o f the stere o
pa i rs will help i n i denti f icati o n of image points and in interpre ta-
ti on o f geo l ogical dat a.** The simple graphical methods discussed
here c ompleme nt the standard photogrammetric t e chniques for geolog ic a l
measurement s a t all s tages o f e ngineering proj e cts .

Example

Figure s 4-15a and 4-15b are photographs of a road cutting taken


wi t h a Pentax Spotmatic camera from two stations, 01 and °2 , The
original foca l lengt h , f = 34.4mm , has been enlarged to 84.5 mm. The
hor i zontal and v erti ca l angles to several recognizable points in the

* T.A. Lang des cribed the us e o f such t echniques in the Sn owy


Mountains project in unpub l ished notes on Rock Mechanics writ t en
in 1965 . He referred to work by W.A.G. Mue ller (1959) in th e
Aus t ral i an Surveyor, Vo l . 17 , No.7.

** Williams p o i nts out tha.t one must be careful i n the field to


i den tif y whi c h photo is t he one t o which the control observa-
t ions apply .
120 Exploration of Rock Conditions

(b)

Figure 4-15. (a), (b) Photographs of a road cut taken from the ends of a surveyed
base line. Horizontal and vertical angles to a number of points in each photo were
measured with a Brunton compass when the photos were taken. The enlarged focal
length (ef) is 84.5 mm. The map of figure 4-15c was made from enlargements to
248.5 mm. (c) Plan and cross section of the cut slope shown in figure 4-15a and b
by triangu lation from the two photographs.

photo field were read with a Brunton compass mounted on a tripod,*


and the distance and direction between 01 and 02 was measured with

* A transit would be preferable, if available, as the quality of


control data limits the map.
Exploration of Rock Conditions 121

(c) SECTION AB PLAN


Half Sco le

B
'-:'0
+

~'j

Q .' .

/N
Scole of Sec llon
0 50 100 fl
I I I
0 100 200 ft
Sco le o f Pi on

(Elevollons In feet ore shOwn In ( ) )

po7
G ,m ero /
comerojpolnl 2
OO lum E levollon
0, 02

a s t eel t ape. A n umb er of key points, rec ogniz ab le in e a ch pho t o ,


we re l abe ll ed on i n di vi dua l t r a c ings ; Usi n g t he gnomonic net
c en tered on t he p rincipal po in t , wh i ch was fo un d from in te rs ecting
diagonals , the tra ci n gs were or iente d as pre vi ous l y s h own (f igu re
4- 1 4 ) ; t h e n hori z on tal an gles t o all the marked poin ts we r e re a d,
a nd t he pl a n (fi gure 4-l5c ) was developed by t riangu l at i on f rom 01
and °2 , Ne x t , ve rti ca l a ng le s o n photo 2 were r ead f or a ll the
lab e l led p o in ts, and t he d ist an ces from 02 to each po i nt were sca le d
from t he p lan, y ielding re lative eleva tions . F ina l ly, geo logy was
i nterp re te d a n d s ke t c he d be twe en the plot ted poi nts.

GEOPHYSICAL METHODS

Geophysica l surve ys conducte d fr om t he g r o und surface can supple -


me nt t he data of g eo lo g ical ma pp i ng . A number of methods of geo-
p hysic al e xp l orat ion are d iscuss ed i n books b y Gran t and Wes t (1965),
122 Exploration of Rock Conditions

Dobrin (1960), Heiland (1946), Jakosky (1950), Van Nostrand and Cook
(1966) and others. Many of these methods have application in
geological engineering as summarized by Griffiths and King (1965).
Here it will be sufficient to list those techniques which help to
describe geometrical and mechanical properties of discontinuities
in the rock.
Geophysical surveys usually yield two types of results. First,
they evaluate a physical property at different points within the
rock mass -- properties like the velocity of sound waves, magnetic
susceptibility, and density. This result may be a direct product of
the instrument readings, but more usually it is calculated from the
data. Secondly, the geophysical exploration will usually reveal the
distribution of the measured quantities over the map; interpreta-
tion of these data, based upon a suitable model of the geologic
structure, will often permit the geometric constants of the sub-
surface structural model to be determined. For example, data might
be interpreted to yield the depth of each layer in a multi-layered
configuration, or the width of a fault zone in a model with a tabular
fault between continuous walls.
The physical quantities measured by geophysical methods may be
of direct value in a design problem. The longitudinal wave velocity,
for example, :is used in assessing potential damage from blasting
operations. But more frequently, the real value in knowing the
measured quantity lies in its associations, such as with the rippa-
bility of rocks; rocks with sonic velocity less than 7000 feet per
second are usually excavatable by ripping, whereas rocks with
velocities greater than 10,000 feet per second will usually require
blasting (Caterpillar Tractor Co., 1966). Figure 4-16 shows how
the sonic velocity correlates with the characteristics .of weathered
granite, including the percent of "core stones" and the porosity
(Polak, 1963). The next section considers briefly some methods which
are relevant to the problem of mapping and assessing discontinuities
in a rock mass.

Seismic Methods

Several geophysical methods utilize the signature and relative


Exploration of Rock Conditions 123

c~a!i!( zone 4

Figure 4-16. The variation of sonic


After Rux ton and % C ore stones % porosity veloc ity in t he weathering profil e of
Berry ( 1957) granite. From Po lak ( 1963).

ar ri val t i mes of elastic waves travelli ng between a shallow rl sho t


point rl and on e or more transducers. If a n u n de rl yin g r ock or so i l
layer has a h igher sonic velocity than t he overlyi ng layer, it will
serve as a refra c ti n g surf ace. In a specific range o f di stances f rom
the s h ot point, t he time ver s u s d is tance data of f irst a rr i va ls f rom
the s h ot will rep rese nt waves travel lin g alon g th i s sur fa c e. Within
this range, the rate of chan g e o f d i st a nce with fi rst ar rival time,
ax/at, is equal to t h e longi tudinal wave velocity in t he r e fracti ng
layer, if it is horizon tal and regula r. If it is i rregu lar or
inclined, a simple graphical interpretation c an r e ve a l t he v eloci ty,
as well as the topography an d depths, as shown in fi gure 4-1 7 . When
t he refracting sur fa ce is a st r at igr aphi c ho r i z on, a f aul t is indi -
cated by a sudden topograp h ic change al on g it, as in fig ure 4- 17 .
Joints an d sh ear zones will rare ly be s een in r ef racti on pro f ile s ,
but they may be recogni zable in con ti nuo us overwa ter r e flection
profiles, using a b o at to advan c e a tran s ducer a nd a r epeating s ho ck
sou rce, such as a rl s parkerf1 (Fanshawe and Watkins, 1 971) . Oth er
meth ods of seismic measu r ement a re somet imes u sed in engin e ering
investi g ations. Cross hole shootin g , with se ismic so u rce in a
borehole and receivers in oth e r bo rehol es, is us e d to map the veloc ity
variations in the subs urfac e . Faults an d c lose join ting will reduce
the n et travel time an d yield app ar ent ly l ow ve loci t ies . Wave fr ont
diagrams, constructed from data wi th t he shot p o int in a dril l ho le
and the transducers in a l in e o n t he s ur f a ce, wi l l s uppl e me nt data
124 Exploration of Rock Conditions

DATA FROM REVERSED PROFILES


"-x.....
~~
tl--------------~ --

'lii
tI:
.E
(I)

E
~ O~------~~---~~------~~------~------~~---~~------~~~
Shot point
I

DETERMINATION OF VELOCITY
ALONG REFRACTOR (V2)
10
uQ)

E5 Vz =6900 m,lsec
~ O~------------------~~----~-----------------------~
I

..... -5

-10

uQ)

E 15 DETERMINATION OF FUNCTION F. ttl + t2- T ) ALONG PROFILE

~
':""N 10

£5
"
~ O~------------------------------~---------------------------------------~
Distance

O~---------------------------------------------------------r---------------'
GEOLOGICAL PROFILE ?
'2
E
4

a.
(I)
6
o

Figure 4-17. Approximate graphical method for locating faults from seismic
refraction "reversed" profiles. Based upon Griffiths and King (1965)

from surface profiles, especially where a lower velocity layer lies


below a higher velocity layer (Meissner, 1961). If the whole signa-
ture of the wave arrivals are recorded rather than just ~he first
Exploration of Rock Conditions 125

a rr i va ls, a ddit io n al info r mation about discon ti nuities in t he r ock


c an be d etermi ne d f rom the wave attenuat ion, as discussed by Polak
( 1 9 71).

Resistivity Surveys

Since mos t rocks are t hemselve s nonconduct i ve , the e lectrica l


res ist ivity of a r o ck de rives mainl y from sa l i n i ty i n t h e groun d
water occupy i ng pore s and f ractures . According l y , rock formation s
wil l dif f er in res is tivi t y becaus e of porosity and jointing
di f f e rences. F ault s t hat act a s contacts may , t here f ore , be mappab l e
b y res i stivit y. In f aul ts a n d shears , t he water content may be
h i ghe r t h an in t he country rock , and anomalousl y l ow r es istivi t y wil l
be measur ed . Oc c a siona l l y , in porous coun try rock , a frac tur e wi l l
a c t as a drain a nd appe ar as an anomaly of high res isti vi t y (St ah l,
19 73 ). Fie ld measurement may be accomplished b y t rave r s ing alon g a
l in e wit h a set of f our e lect ro de s a t fi xed spacing, ("Fixed We nn er"
arr ay). Ba tte ries or an AC c urrent source are conn ec ted t o s e t up
a c u rre n t f l ow ( I) bet ween t h e o ut e r e l e ctrodes, whil e t he po t en t i a l
d rop (6V ) between the i n ner elect rodes i s measured with a mi lli -
vol tmet e r. The r e s istivit y ( r) i s calculated from th e re l a tions h i p

r 2'TT a 6V I I (10 )

South North
400r-----r-----r-----r---~~--~~--~r_--~r_--_.----_.~--~

Fault
April 1972 zone
f\~
E
I 6
,',: 'q:'
~

C: , .~: ~

?:'
200 ,p Ji '\
I: .
':; ii
I •
~ , .
/ :
'iii
Q)
Il:
r
0
Q 200 800 1000
Distonce

Figure 4~ 18. Results of a fixed Wenner resistivity profile across a fault zone. From Stahl (1973)
126 Exp loration of Rock Conditions

Figure 4-18 shows t h e resistivity variation in moving fixed Wenner


arrays, wi th e l e ctrode spacings of 10, 2 5 , and 100 feet, across a
resistive fa u l t zone. Inte rp retation of resistivity surveys is
more complicat e d than interpretation of seismic profiles; consequently
one usually checks the interp retations by selected trenching or
boring. (This is a good practise in engineering geophysics work in
general.)

Gravity Surveys

Measureme n t of changes i n the gravity field from point to point


may be used to map changes in geology because density changes between
different f o r mat i ons produce anomalies. The subject is complicated
and has not been applied extensively in engineering practice. Very
precise gravimeters now available will respond to 1 mm changes in
elevation a n d are b ei n g use d to detect vertical r ock movement. An
approach fo r c h aracterizing the ro ck mass porosity in-situ is to
measure the apparent density of a rock mass in which the rock specimen
density is known (Eaton et aI, 196 4). If the mass density of a sample
of the rock is P while the bulk mass density deduced from the gravity
s
survey is Pb' t he add itional porosity of the formation beyond that
represented i n t he specimen is ~n. = 1 - Pb/P. I n saturated rock,
J s
~n. (p
- Pb )/ (p - P ) where P is the mass de nsity of water.
J ssw w
If this poros ity is created by a given pattern and spacing of joints,
it should be p os s ible to dedu ce the average aper tu res of each set;
or if the ape rtures are as s umed, the average spacing of each joint
set can b e calc ul a ted . Interpretation of field gravity measu r ements,
with excellent levelling control, will yield density values for
different r o ck bod ies if the true geometric p r operties of each body
are dete r mined by drilling or otherwise.

Magnetic Methods

The int ensi ty of the eart h's magne tic field at a point on the
surface is the sum of two vect ors -- the induced magnetic intensity
derived f r om t h e eart h 's magneti c field acting on rocks of g i ven
magnetic s u scept i bi lity, an d remanen t magneti sm in t h e rocks. Si nce
rocks vary in their magnetic suscepti b i lity and r emanent magnetism,
magnetometer surveys can map the boundaries of formations. Rocks
Exploration of Rock Conditions 127

derive magnet i c susceptibility from the minerals magnetite, hemat it e


and py r rhotite , and as these min e rals are more prevalent in the bas ic
igneous rocks , the magnetic highs correspond to occurrences of such
r o cks . The California Departmen t of Water Resources used aero-
magnetic surveys successful l y t o l ocate bodies of serpentine in
connection with the planning of water tunnels and the Corps o f
Engineers used surface magnetometer surveys to map buried serpentine
bodies at a dam site. Fracturing and weathering of rocks demagne-
tizes them, so one can expect faults and highly jointed igneous
r o ck masses t o show up as anomalies of low magneti c intensity.
Wiebenga and Polak (1969) used a "micro-magnetic" method to revea l
joint trends in an area of rock thinly masked by weathered debris.
An area of about 100 feet by 100 feet was covered with a grid of
measuring poin t s on ten foot centers.
Unfortunately, it i s often easier to make good geophysica l
measurements than to provide good interpretations. However, geo-
physica l techniques are generally relatively inexpensive and can
yie l d u seful information under the proper circumstances.

DRilL HOLES

Geologica l field mapping rarely yields sufficiently detailed


data for design because the surface exposures give incomplete details
of t h e stratigraphy and structure. Even i n the planning stage , it
is usua l to invest a significant sum in bull-dozer cuts or trenches
and some dril l holes to determine definitive answers to crucia l
questions centering on the geology. Geophysical measurements, aeria l
photographic interpreta t ion~ and shallow excavations and test pit s
must e ventual l y be supported by dri l l holes, and sometimes,
e xploratory t unnels, in which tests and observations can be made and
from which samples of the rock can be studied. We will consider
meth o ds o f dril li ng and sampling rock and the important question o f
orienting the core so that the attitudes of discontinuities can be
measured .
Dril l holes are expensive, c.a. $25.00 per foot fo r diamond
drilling , with pressure testing , boxlng of samples, and logging of
t he hole and core. The amount of drilling can be minimized if
128 Exploration of Rock Conditions

planning of each hole considers previous geological knowledge,


including previous drill holes. Evaluation of the adequacy of a
drilling program requires appreciation of the probabilities of
intersecting the possible targets of search with the given pattern
of ,holes. If the sampling of a site consists of a series of parallel
planes such as deep trenches or geophysical profiles, then the
probability of encountering a specified target is approximately the
ratio of the length of the target, normal to the line of the profile,
to the spacing between profiles. When linear samples or observations
are made, as with parallel drill holes on a regular grid, the
probability of encountering a target is related to the ratio of its
area normal to the sample lines to the area of the unit cell of the
exploration grid. When little is known about the geological struc-
ture, the greatest probability of intersecting and finding targets
of search will be obtained from a square grid of holes. Figure 4-19,
from Slichter (1955), gives the probability of encountering a target
of circular or rectangular area normal to holes on a square grid with
spacing S. Given a particular spacing of drill holes, this figure

Qj 1.0r-----oo;;::r......::--~c____._----..._----.,.___----r__---___;----......,

~
~
....
.2
~ 0.8
c: Disc
o
U L/W: rectangles
~
2
5 06
u 0 .
If)

6~
o~
<l
g'-
t; °0.4
Cl)

~
2c:
....
'0 0 .2

.c
o
.0 O L __ _ _- L_ _ _ _~________L __ _ _ _ _ _~_ _ _ _ _ _ _ _~_ _ _ _ _ _~_ _ _ _ _ _ _ _~

~ o 1.0 2.0 3.0


a..
Relative Grid Spacing, SIlO

Figure 4-19. The probability of a single intersection of disc or rectangular target


using a square grid of drill holes with spacing s; after Slichter (1955).
Exploration of Rock Conditions 129

wi ll p e r mi t you to det ermi n e the degree o f c e rt ai n ty wit h which you


can be assu re d of having found a g eolog i c al f eat u re , such a s a fa u lt,
a hydro thermally al t ered z one, a cavi ty, a buried valley, e tc. F or
example, with rectangular targe t s o f l ength to width rat i o l /w eq ua l
t o 2 , and a grid spac i ng S equal to 2(lw)i , the probabi lity of
e n countering t h e target is onl y 25%. A dri ll ing program c an not be
expe c te d t o find t arget s a ppre ci ably sma l l e r t han the spacing b et wee n
dril l ho l es .

Methods of Drilling

There a re n umerous ways to dr ill rock , e ach devel ope d f or a


s p ecial comb i nat i o n o f f act ors. F or t h e s mall d i amete r e xplorat ion
holes o f t h e en gine e ring geo logist, o nly cor ing metho ds , mainly
di amond dr i lli n g , pro vide a s at i sf actory rock sample; thi s t hen i s
t he prin c ipa l a p proach u se d t o e xp l ore det ai l s of t he sub sur face
s t ructure. Less e xp e nsive non -coring methods can satisf y ot he r
expl o rati on ne eds, f or e x a mple : to p r ovi de relativ e ly cheap acce ss
to a p o i n t where corin g is to b eg i n ; to e mp lace se ismomet e r s , p iezo -
me ters , tilt -mete rs a nd ot her instruments; to me as ure wate r l e vels
or rock t empe ratures ; to map t he t op of rock su rface ; or t o p rovide
blas t ho l es fo r se is mic e xpl o r ati on.
Non- cor ing drilling is performed by pe rc ussive or she a ring
me t hods i ndivi du a l ly or i n combin at i on ( Table 4-3) . So il exp l orat i on
met hods by a lte r na te dr ive samp l in g a n d wash boring , combi nin g a
chopp ing ac tion wit h e rosion b y wat e r, are n o t often app r opri ate
f or rock s because t h e "sp li t spoon" can not be driven wh i le the
wash bori n g pro duces a v al uele ss sampl e and a damage d hole . Dri lli ng
by raising and lowering a ch i s el (churn dr il ling) is no l onge r as
e c onomical as p n e umat i ca lly d ri ven percussion dril l s, whi ch a re
driven fr om t h e t op or by down-t he - ho l e hamme rs . Churn drill ing i s
sti ll use d in t he ab sence of compresse d a ir. In so ft rocks , a u ge ri ng
and combined p ercussion-auger drills c a n be u s e d , v e r y c h eap l y.
Howeve r, ro t ary dr il l ing with c arbi de blade s ( dra g b its ) is f ast er,
even in rocks of medi um h a r dn ess ( e . g . coal , some sha les a n d sand-
stones 1 evapor ites ,. marls a n d some serpen t in es a nd limestone s ).
Rotary dr i l l i n g us i ng toot hed ro l l i n g c ones ( "t ricone " bit s ) can
130 Exploration of Rock Conditions

TABLE 4-3
Drilling Methods

Data from McGregor (1967) and other sources

Augering - Disturbed samples of soils and soil-like rocks; 2 to 241r


diameter holes, maximum depth 60 to 100 feet; holes usually
vertical.

Wash Boring -- Used for making a hole, without return of a sample, in


soils and in soft rocks; only in vertical holes.

Center Sample Rotary Drilling (Becker CSR) - Chips or cores of all


rocks ; holes 3!1r and larger, up to 750 feet deep; holes may
be inclined up to 45 degrees

Churn Drilling - Used for making a hole, without return of a sample,


in soft and medium hard rocks. Holes commonly 2" to l5 1r in
diameter, up to 4000 feet deep; usually vertical.

Track or Wagon-Mounted Percussive Drilling - Used for making a hole,


without a sample, in hard rocks. Holes commonly 1-3/4" to
4-1/2" diameter, up to 100 feet deep, in any orientation.

Down the Hole Percussive Drilling - For making holes 4" to 6" in
diameter without a sample, in hard rocks. Usual depth range
120-200 feet. Can be drilled in any orientation.

Rotary Drag Bit Drilling - Very fast drilling in soft to medium hard
rocks, but without a sample. 2-3/8" to 7-1/8" diameter holes
in any orientation. Maximum depth in the range 100 to 250
feet.

Rotary Tricone Drilling - Small diameter core sample obtained in


the larger hole sizes in all but the very hardest rocks.
Holes up to 3" to 12" in diameter to any depth; rigs for
drilling up to 200 feet in any orientation are common.

Shot Drilling - A continuous core sample in all but the hardest


rocks with holes up to 6 feet in diameter; rather slow. Small
diameter shot drilled holes have been drilled up to 1000
feet deep vertically. Calyx holes rarely are drilled deeper
than 100 feet.

Diamond Drilling - Continuous core sample of all rocks except highly


fractured rocks of great hardness. Common hole diameters
15/16" up to 7-3/4". Maximum depths of rigs are 200 - 1500
feet. Wireline equipment can be used in steep holes.

penetrate all but the hardest rocks. This method was developed by
the petroleum industry where high bit pressures can be provided by
Exploration of Rock Conditions 131

the weight of the drill string. For shallow drilling, very heavy
rigs are needed. A continuous rock core is obtained using diamond
coring bits (figure 4-20) or shot drilling methods. The latter work
by turning a slotted steel cylinder on continuously supplied abrasive
steel shot; this slow expensive method is used only on very large
holes or for smaller holes in situations where diamond bits are
unobtainable. Czech engineers, for example, have used shot drilling
to explore the foundations of dam sites because they could not pur-
chase diamond bits. Large diameter shot drilled holes (36" to 60"
"calyx" holes) were used by the T. V. A., the Corps of Engineers, and
other engineering agencies to permit downhole inspection of the rock
by a geologist.* Oriented core techniques, and to a lesser extent
bore-hole cameras and television devices, have superseded calyx drill-
ing for exploration in all but very special circumstances.
Good drilling practice reduces costs and is vital in coring for
civil engineering work since the quality and completeness of the
sample are sensitive to drilling efficiency. Thrust, rotation and
flushing action must be properly adjusted according to equipment
and rock conditions: Thrust applied to the bit forces the mechanical
work to be done on the rock and not on the drill. In diamond drilling
it is the force per carat which controls; too high a value breaks
diamonds and too Iowa value polishes them. Flushing the bottom with
fluid, water or compressed air, transports the cuttings from the rock
breakage region and cools the cutting tools; an insufficient flush-
ing rate risks plugging of the annulus, indirectly causing blocking
of the drill core. The speed of rotation affects the cutting action
of the bit; it varies according to the bit design and rock type. The
optimum speed of rotation (in the range 250 to 1500 rpm) is deter-
mined by the minimum thrust. Since the controlling variable is the
tangential velocity of the diamonds, larger diameter bits must be
turned more slowly than smaller ones.
For maximum core recovery, diamond drills for civil engineering
geology should be "hydraulic feed" types, in which rotation speed

* Calyx hole inspection is unacceptably hazardous unless an air line


and a protective man car are provided.
Goodman-Geologica l Eng.-lO
132 Exploration of Rock Conditions

Wo ter Sw i vel

i,·rPum p

Dri ve Pipe
c D ~ c:::::, o v D
: -

Cas ing _- =- _ _-

H - 15 Dri ll
Co lumn Moun ted

Dou ble Tube


Core Barrel

Ret urn Water


Hose to Sump

Figure 4-20. Typical diamond drilling set-ups; (courtesy Joy Manufacturing Co.) .
Exploration of Rock Conditions 133

an d bit p re ss u r e can be var ie d independently . Large di ameter h ard


ro ck dr i llin g re qu ires r e latively sl ow rotation with h igh pressure
wh e re as f r iable s o ft rocks want a mo der a te speed of ro ta tion, with
low p r e s s u re. The h y d rau lic feed is u s u a lly helpful in layered
format ion s of v ar iabl e h ardn ess, where h igh core loss will result
f rom impro pe rl y contro lled d ri l ling owi ng to the break ability of t h e
in d ivi du a l laye r s .

Core Barrels

S i zes of diamon d d r ill holes were st a n dardized in 19 30 by t he


Diamond Core Dr ill Manufact u r e r s As sociation (X series) * . Later,
t he nee d for lar ge r d iameter ho l es in g eo logical exploration for
e ng i n eeri n g works led to a "larg e diame t er se rie s" (Tab le 4-4 ) . The
c o re bi t and c a s in g sizes i n standardized series desi gn ations are
"nes t e d" (Tab le 4- 4 ); i.e, i f dr i l lin g b eg i n s with a n NX cas i n g bit,
(3.6 i nch diameter ho le ), a nd at some po int steel casi ng is installe d
to preven t c aving o f wea th ered or s eamy material, drilli ng can con -
t inue wi th a n NX bit ( 3 inch h o le) which admits BX casing a n d so on .
Me t ri c equipment does not cor r e sp on d in size to th e X se ri es.
La r ge di ame t er c o r e barre l s wi t h sop h ist ic ate d desi gn feat u r es
a re us ually re q u ired i n civil eng ine e r i n g e xp lo rati on to preven t
co r e l oss or d amage . Co re l oss arises from g ri ndin g , bre akage an d
eros i on . Wit h a s in gle tube core b arre l, e sse ntially a se ction of
casin g fo ll ow ing t he bit , th e wash water travels alongs i d e t h e core
a nd di scha rge s at t h e bit, retur n in g i n the an nulus out s ide the
b a rre l . Rock c h i p s erode d by the wash wa ter c an jam the bi t, blo c ki n g
fr ee e n t r a nce of new c o re i n t o the barrel and leading to core break-
age. Eros ion of fri a b le material s se r i ously damages the sample and
washi ng out of s eams de st roy s t h e con tinu ity of the samp le, e xp osin g
it to bre akag e inside t he ba rre l. The d oub l e tube core b a rrel wa s
desi gn ed t o is o late t he core from t he wat er stre am (figure 4-2 0 );
wat e r flows down between th e i nn er and o u t er barr els, discharges

* The Ca na dian Diamo n d Dri l l i n g As s oci a ti o n st a ndardize d s i mi larly


in 1950 (XT se ri es ) ; in 1956 the coupli ng an d t hread details
were modifi e d i n a n ew st a nd ard ( W s eries) but diameters of h o les
of t h e X se rie s g i ve n in Ta bl e 4 - 4 were retained.
134 Exploration of Rock Conditions

TABLE 4-4
Standard Sizes of Drill Holes
and Typical Sizes of Drill Cores

TYPICAL EUROPEAN SIZES

Hole Diameter Core Diameter (mm)


(mm) thick walled bits thin walled bits

146 120 Swedish Di amond


131 105 Rock Drilling
116 90 Co.
101 75

86 58 72 Craelius Co.
76 48 62
66 38 52
56 34 42

U.S. AND CANADIAN SIZES

Name Hole diameter Usual core


(in) diameter (in)

6 x 7-3/4 7-3/4 6 "Large Series"


5-1/2 x 4 5-1/2 4 double tube
3-7/8 x 2-3/4 3-7/8 2-3/4 swivel barrels.
WX (NX casing bit)* 3-5/8 2-13/16
NX 3 2-1/8
BX casing bit* 3 2- 3/ 16
BX 2-3/8 1-5/8
AX casing bit* 2-3/8 1-3/4
AXT 1-7/8 1-9/32
AX 1-7/8 1-3/16
EX casing bit* 1-7/8 1-3/8
EXT 1-1/2 1-5/16
EX 1-1/2 7/8
1-1/4" BH 1-1/4 3/4
XRT 1-3/16 3/4

WIRELINE CORE BARREL SIZES (Longyear Co. )

Name Hole Diameter Core Diameter


in. mm. in. mm.

PQ 4-53/64 122.6 3-11/32 85.0


HQ 3-25/32 96 2-1/2 63.5
NQ 2-63/64 75.8 1-7/8 47.6
BQ 2-23/64 60.0 1-7/16 36.5
AQ 1-57/64 48 . 0 1-1/16 27.0
Exploration of Rock Conditions 135

TABLE 4-4 (continued)

MINDRILL SERIES S CORE BARRE L SIZES

Name Hole Di ame t e r* Core Diame t er


in. i n.
- --
7S 6.290 4.375
6S 5.31 8 3.750
58 4.347 3 . 063
4S 3.630 2.625
HXS 3 . 906 2.875
NXS 2.980 2.155
BXS 2 . 360 l.65 5

* Dimension given is outside diame ter o f r e a mi ng sh ell .

Figure 4-21. Several types of diamond drill bits; (courtesy of Mindri ll, Ltd.).
136 Exploration of Rock Conditions

Figure 4-22. Ground ends of core.

against the core inside the bit and returns outside of the outer
barrel. In erodible or seamy formations, the water stream can be
further separated from the core by using bottom (face) discharge bits
(figure 4-21). Longyear Co's "L" series core barrels (e.g. NXL) and
Joy, and Sprague and Henwood's "M" series (e.g. NXM) reflect improved
designs which among other things discharge the wash water closer to
the bit.
Blocking and wedging of core inside the barrel can cause core
to break, especially if it is of small diameter. Ideally, core
barrels should not be smaller than NX size to explore discontinuous
rocks for civil works.
Grinding in the barrel occurs when previously drilled core
turns against fixed core still attached to the bottom (figure 4-22).
Many feet of core can be lost in this way. Grinding can be minimized
by wi thdrawing ("pulling") core frequent ly so that the barrel is
never allowed to fill up. However, frequent interruption of "drilling
and withdrawal of all the rods to empty the barrel proves costly.
In vertical and in steeply inclined holes, "wire line" equipment can
be used (e.g. Longyear's Q series) in which the inner tube can be
uncoupled from the core barrel and pulled to the surface by a wire
Exploration of Rock Conditions 137

cable in side t he string of special hollow dril l rods. While the t ube
is be i ng e mpt i ed , a seco nd inn e r t ube is lowered and drilling con-
ti nues. If core separates along natural or new f ractures i t can
wedge against the sides o f t he barrel , blocking e ntrance of new core.
With Mindrill's S barrels and the L series barre ls, c irculation of
wat e r c uts o f f as soon as a block occurs t o warn the driller. To
pre ven t grinding, t he core should be prevented from rotating by
means o f a free swivel joint betwe en the top of the inner tube and
t he he ad o f the barrel. Ball bearing swivel type heads are now
general l y avail a bl e for al l siz e s of double tube core barrels, e . g.
t he L , M and "Large " series mentioned above , and should be selected
f or g e o l ogical engineering.
I deally, core should be removed f rom th e sample tube (inner
tube ) without disturbance to preserve mating across natural f rac t ures
a nd to preven t loss of f illing ma teria l or small rock pieces.
Longyear , Mindr i ll and other manu f acturers provide a core barr e l
eit he r with a split inner t ube or as an additional t ube sample holder
wi t hin t he inner tub e ( fig u r e 4-23a), wherein the spli t t ube is
e xt r uded hydraulically from the inn e r tube and separated t o e xpose
t h e core wi t ho ut disturbance (figure 4-23b). A t riple tube barrel
can be made b y i nserting a spli t plasti c tub e inside the inner core
barre l of a double tube t ype .
Good core recovery can be difficult to achieve in friable
mate rials . Christensen Diamond Products Co. develope d a rubber
s l ee v e core barre l which encases the core in a tigh t fitting neopre n e
sl eeve as it come s into the barrel. This is par t icularly useful
when drilling f riab l e rocks with drilling mud which can invade and
dama ge the core. Another special barrel i s Chris t ensen's "pressure
core barrel ", which tries to preserve the original por e f luid in
the core sample by seal i ng t he i nner tube as soon as it is occupied.
Good core recovery is also difficult to achieve in materials of
variable q u al ity , e. g. wh e re hard rock and soil occur in the wea t h e r-
ing profile. F or uniformly soft intervals, large diameter design
cor e barre l s ' have been equipped wi th soi l sampling extensions on
the i nne r t ube which work l ike drive sampl e rs when pushed ahead of
t he coring bit . Be rents ( 1961 ) reported a design by the Snowy
138 Exploration of Rock Conditions

Sa II check va Ive pro-


tects the core from
circulating fluid and
Bottom discharge bit
releases pressure to
with step type lifter
perm it entry of core
Duo I shu toff
valve signals core
block. Easily adjust-
ed fordrillinghardor
soft formations

Threaded spindle for


accu rate adj u st ment
of gap between lower
end of Inner tube as-
sembly and bit

(a)

Figure 4-23. (a) Longyear triple tube wireline core barrel (03). (b) Extruding the split inner tube of the 03
barrel; (courtesy Longyear Co.).

Mountain Hyd.roelectric Authority* for use in decomposed granite


containing layers too hard for the soil sampling extension on the
NM barrel. A spring was placed between the back of the inner tube
and the drill string; when fully extended, the soil sampling exten-
sion protruded beyond the bit; but a thrust of four hundred pounds,
the minimum for drilling soft weathered granite in this case, would
compress the spring sufficiently to expose the diamond bit.
It can be appreciated from the above that the sampling of rock
for exploration of structural features demands careful use of
precision equipment. It is unlikely that the available drilling
contractors at a remote site will have access to the most advanced

* In conjunction with Triefus Industries, Australia


Exploration of Rock Conditions 139

tools, such as the Mindrill S series, Longyear Tripl e Tu be (V 3 an d


Q3) and comparable p r oducts, nor be expe rie nced in t h eir u se.
It is
a good idea for an engineering organiz a tion to purchase its own core
barrel and to become proficient in operat in g it. Some supplie r s of
good core barrels are: Acker; Atlas-Copco; Boyles Br os.; Chri st e nsen;
Joy; Lon gyear; Mindrill ; and Spragu e an d Henwood.

Logging of D rill Holes and Core Sam ples

The drilling l o g, made by the dr illing i ns p e c tor or the dr i l le r


himself, indicates the k ind of equi pment us ed at each st e p, _ the rate
of drilling, the lo c a tion and orientation of the b ol e , t he dep th of
water loss, caving, exc essive vibrati on s, b lockages a n d ot he r i n c i-
dents, as well as any other technical details re l a te d to t he work .
The geologist's log, on the ot h er h and, sh ou ld gi ve comp l ete,
descripti v e details a b o u t the core and observations an d measure me nt s
in the borehole, e.g. water pressure tests. Ide a lly, t h e logs are
made routinely by a geologist as he opens the core barre l ; t h i s is
especially important i n shales and seamy ro c k s wh ich ch ange a ppear-
ance and character markedly when the core dries out. I n s u ch ro cks,
an effort should be made to preserve the field moist ure conten t b y
sealing the lengths of co r e, or at least a r epresen tative se l e c t ion,
in plastic bag s. The core should be place d in sturdy woo den boxes
with hinged cove r s and divided in c omp ar tments of the s ame le n gt h
as the missing sections and the enti re box photog raph e d ro utinel y
before being placed in storage. A ten dency e x ists to e c o nomize on
storage requirements by closing up gaps in t he c ore r e co rd ;
photographs of such core are misleading to u sers , as is a log with
broken lines to shorten monotonous sectio n . The co re bo xe s s hould
be stored on frames allowing any box to be remove d witho ut fi r st
displacing those on top, and a cu r r ent in dex map to t h e core s h e d
should be displaye d.
The excellent logging procedu re used by the Snowy Mounta i n .
Hydro-Electric Authority was described by Moy e ( 1967 ). A c olumn ar
section was made down the center of eac h sheet and c orrespondin g
graphic response logs were drawn to show the number of f r a c t u r e s per
foot (fracture log), the percent core re cove ry , and t h e wate r l oss
140 Exploration of Rock Conditions

"""'5 (Oct, 1m)



Hol. , __0_-_4.:.....-_ _ Boa I ___'_7_ __

CORl: SKrrCH

I' ,,
I I I I , I i I I I I I I I I J:tl~ "!'wI !. 11f I I I I I I I Ii, I I I I I I I I r
GiY@ (i)
M ~ _=- ~~J_ ___ ~
,
u
IY '
- ~ 0- -----n--if2-
LL-L _ _
__ _
- w- ~ - ---, --- ~~-.
~_ ~'- _ _ ---1 _ _ _L- ___
A @(f) tv( M

CD. ® .• tc • Joint. i 14 • Mech&nle&l Bre&kj • - ep..eer

Figure 4·24. Format for geological


log of diamond drilling introduced
by Prof. Tor Brekke; the log was
prepared by Mr. Larry Myer;
(courtesy of the Pacific Gas and
Electric Company, San Francisco).

in pressure tests. Considerable space was provided to describe


petrology, weathering, and discontinuities.
A somewhat different format for presenting the results of ex-
ploratory core drilling was developed by Professor Tor Brekke of the
University of California for the Pacific Gas and Electric Company
and is reproduced in figure 4-24. Each core box is represented by
two facing pages in the log book: On the bottom left is a photo-
Explora tion of Rock Conditions 141

76-545 (oct, 10/73 )


A
HEU·~ PlJ}·~ED STC l\.~GS F.r.(_:'..~I: T
FIELD CORS L00

Hole # _--"0_-..:..
4 _ _ _ _ _ Box # _ _1,--7 _
DEPrn INTE.~VAL OF BOX : 23 5 .6 to 250 . 4 INFOEHATIOr< fROM DRILLERS' LOG
CORE RUNS TNCwD'ED; - "'2.i4T t o 2'44:'"5 Drill ing Re.t~ : .1- .4 ft/min
244 . 5 to 2s4:'5 We. ter Lo ss:.....;;;,.l0;",;0'-;1_ _ _ __ -,--,-_
- - - to - - - Remarks: '...'" t " r r et urn a[ end of
core __n_ _ _ _ _ _ _ _ _ __
r u~

CORE RECOVERY: _ _ -100% _ _ _ _ __ _ __

ROCK
ROCk. type, color, d egr':E of w6athering: Fre s h ~re·. granod ior it e

-------- -~----------.-.--- ---------

Texture, structure/foUati on: Med ium g rained. Fal i a t i o n indi s tinct a t /.4 - 50°

CHARACTER OF JOItfTS #1 #2 #3 #~
Inclination: e
59° 30° Il J5°

Rough Smu o l h -) f'llh.' th Kough


Shape & RougP_'1e ss:
~ ub plan ar Pl an"r Pl Anar Suhp lanar

S ur f ace We atheri ng: Fresh Tr ace Fresh Trace to


(rust) Light
(rust)
Coati ng8:
C.:vln Crf' y I~ lc l.hl· r l l l C Light Coal
CoaL or C-c ce n Yellow Clay
(,.!l c (oat
Filling Material:

Remark!! :
#1 Subparall e l foli a e ion

#2 Per pe nd icul ar foliation

#3 (3) Exhib i close ioin r s pa c i ng even t huu g h a l mo s t axial in orien ta t i o n

#4 Or ie n t a t ion v a des 2:,;:0_°--;-,:.;35'-o_ _---:-7_~---;----;----------­


( 4 a) - has light to m0gera r e we athering (Cl I ~t)

OTHER DI SCONrINU ITIE S


CA) - Rr oken zone dut': t o clo se jOi ntin l" s urfJc,,~ n :' pieces r h l o rllic

(B) - Broken zo ne due co Ll ose j o , ntin h .Il hl ,"""-/o-,n\ , ,, l ·· rt?,\k"g~

Da t I!! Logg ed _ _-=B:.J.,/..:.-7,-


1 ,-
7 4'--..

g r aph of the c o re box, pre fe rably in c o l o r; above it is an an no tated


sket ch o f th e c o re as s e en in t h e phot o g r ap h. The symb ols used in
the c ore ske tch a re fully described on th e righ t page togeth er wit h
i dentifying i nf ormation and h ighl i ght s of the dril le rs' log. The
mai n joint set s are cl a ssif i ed by or i en t a tion an d cha racter, an d
additi ona l in d i v iduals a r e no t e d u nde r the se~tio n "oth er d iscont i nui -
tie s " . This me t ho d o f logg ing an d d isp laying th e information from
142 Exploration of Rock Conditions

the core box not only presents the data vividly, but provides an
interpretation and classification of the discontinuities that will
be appreciated when entering into analysis and decision making.
The potential penalties for careless or faulty logging proce-
dures are high, as the logs are often included as legal entities
available to bidders and become a part of the contract. One legal
case concerning rock excavation centered around a marginal notation
"probably needs blasting" scribbled on a log sheet in the field.
In Zambia, a construction company blamed a faulty log of a boring
at the site of an underground power house for a chain of events which
led ultimately to the forfeiture of the contract and the bankruptcy
of the company*; in this case it was alleged that biotite schist
bands within a biotite gneiss, logged uniformly as "sound gneiss",
created a falling rock condition.

ABSOLUTE ORIENTATION OF STRUCTURAL FEATURES IN DRI LL CORE

A drill core is considerably more valuable if the true attitudes


of planar discontinuities can be logged. The orientation of a
discontinuity is often its most immediately significant attribute;
furthermore, the use of the sub-surface data for correlating struc-
tures from hole to hole is greatly enhanced when the core is absolute-
ly oriented. When a distinctive discontinuity is recognized in the
core of non-parallel drill holes, one can calculate its orientation.
In the majority of cases where this can not be accomplished, absolute
orientation can be determined by reference to other structures in
the core whose orientation is known, either through knowledge of
the regional geology, or from pointwise down hole orientation
measurements. Such measurements can be made using: the Craelius
core orienter; bore hole periscopes, cameras, or television devices;
paint marks on the hole bottom; or paleomagnetic determinations. It
is also possible to perform continuous orientation of the entire
drill core by scribing it with knives in the core barrel by using

* "The Story of Borehole 4, Kariba North", New Civil Engineer,


8th February, 1973, pp 10,11.
Exploration of Rock Conditions 143

t he "integra l sampling t echnique" and by down hol e "dip me te r' !


surveys .

Core Orientation by Reference to Known Structural Attitudes

Rose n gre n ( 1970) , s howe d how le ngths of core can be p lace d 1n


a consis t ent relati ve orientati on. Though tedious , thi s can b e
r ewarding if core re c ove r y has been excellen t. The core p ie ce s are
la id on a "V " t rough in proper sequence and rotated as neces s ary to
f it a l l t he mating pi eces toge ther . The entire length of core i s
t hus divided int o a s e ries o f continuous sub-lengths; a cont i nuo us
refere nce line is painted down each such sub-l e ngth, and t h e points
where i t must b e interrupted a r e clearly ma rke d. Planar featu res
wil l intersect the core to form e llipses (figure 4-25a). The angle,
Q, between the core axis and the major axis o f an ellip s e of i nt e r -
section can be measured wi th a cant act gon i ome ter (figure 4 -25c).

~~~
(a)

II 90-0 # I

YPlanep (b)

I f3 plane Co re sect ion ]


,'- -- ---- - -- --- -- -'-:~ ---//~ - ~ ----: ~,-,-:
./ .,./ / ',;' ~
I

,
I

- Core gu id e \
I
:
/ - ............ ..... ~
/'
':"
~------.....-: -:::.'
Reference
line
CD
Drilling
direc t ion - --- ............
"
+
- Gra dua ted circle
(c) Fr i c tion kno b

Figure 4-25. (a) Rosengren's reference line and the angles a and (3. (b) A simple
method of measuring (3; {1 = 2500 (courtesy of Golder and Brawner Associates,
Vancouver). (c) A goniometer for measurement of a ; (courtesy of Golder and
Brawner Associates, Vancouver).
144 Exploration of Rock Conditions

Figure 4-26. Core being referenced and described; note the continuous reference
line on the mated lengths of core laid in /Iv" troughs (courtesy of Dr. K. Matthews,
Mt. Isa Mining Co., Australia).

The downward end of the major axis of the fracture ellipse meets the
core circumference at an angle S from the reference line (measured
clockwise looking in the direction of drilling). This can be
measured routinely using a circumferential band calibrated in degrees
(figure 4-25b). Figure 4-26 shows core being referenced as above.
The angles a and S determine the orientations of any fracture in
coordinates local to the particular reference line. If the absolute
orientation is known or can be determined for any planar feature
(other than one perpendicular to the drill hole), all other planes
can be oriented absolutely.
The stereographic projection facilitates the solution. Consider
the ellipse formed by the intersection of an inclined plane of
discontinuity and a vertical cylinder (figure 4-25a). The minor
axis of the ellipse -- the strike of the discontinuity when the core
is vertical -- is the line of intersection of the plane of discon-
tinuity and the plane perpendicular to the axis of the core. Since
the latter is the plane in which S is measured, we shall call it
"the S plane". The major axis of the ellipse of intersection projects
into the B plane at 90 0 from the minor axis (the B value for a plane
is measured from the reference line to the major axis). Respecting
Exploration of Rock Conditions 145

the sign convention given , the major axis wil l be in the lower hemi-
sphere if t he hole is downward. Figure 4-27a shows a s tereographic
pr oje ction relating the angles SR and a
R for a reference plane (R)
o f kn ow n attitude in a drill hole (H) of known orientation. The S
p l an e is constructed as the p l ane normal to H. and its intersection
wit h R, together with t he measured va l ue of SR determines the position
of S = O. In this i llustration , t he drill hole H is inclined 70° to
0
t he n o rth, while the reference plane strikes N40E and dips 40 to
t he southeas t. SR = 7 o and u 0
60 .
R
Now that the orientation of the reference line in the 8 plane
has b een det e rmined , any other plane defined by values of a and S
may be oriented absolutely. For example, (figure 4-27b ) a plane PI
in t h e same s ection of core as the reference l ine of figure 4-27a,
0
has values a p = 45 and Sp = 65° . First mark the point P' at 65°
from S = 0 along the S plane . Then, in the plane BP', determin e the
posit i on of t he major axi s of the ellipse MAJ(P) located 45 0 f rom
H. Next plot one of th e minor axes of the ellipse MIN(P) at an
angle 90° from P' along the S plane (the sign does not matter). Plane
p is, de f ined by the great circle common to MIN(P) and MAJ(P) whi,ch
0
i n t he example st rikes N75E, and dips 64 t o th e South.
Th i s solution, as wel l as others to be discussed , requires a
knowledge of the orientati o n of the dril l hole. Boreholes , especially
i nclined ones, tend to wander f rom their initial orientation aft e r
a hundred feet or more, depending upon the nature of the rock ; the
problem can be acute when t he angle (a) of foliation or bedding
becomes small . Wherever core orientation is being considered , it is
good prac ti ce t o order a survey for the inclination of the dril l
h ole . This can be done by a variety of methods as discussed by
Cummin g ( 1956) .

Approach Using Non-parallel Holes

If a planar feature ma i ntains constant orientation throughout


a v ol ume of rock , and has a distinctive character making it recog-
nizab l e in different boreholes , it can be orie.nted absolut ely from
k nowl e dge of i ts a values in non-parallel holes. Be dding , cleavage )
schistocity , f oliation, any joint set with a strongl y expressed
146 Exploration of Rock Conditions

H :: orientation of drill hole NORTH


DR:: dip of reference plane
NR :: normal to reference plane
MINlR) = minor axis of ellipse of intersection
MAJ(R) :: major aXIs of ellipse of intersection
Reference
plane (R)

,
J

I
NR-~90-QR I
~--__ ~ I
QR~ I
--0-__
-- ............ -.......... I
............. I
MAJ(R~ ........... j3::0
,.......... 11

/ ............. v~
/
/
e>DR
900 /
'/
/
/
/
/
MIN (A), /' j3 plane

--- -=--~---

(a)

Figure 4-27 _ (a) Absolute orientation of the reference line in a lower hemisphere
stereographic projection. (b) Absolute orientation of plane P given its lX and {3 values.
lXp = 45° and {3p = 65°

statistically preferred orientation, faults, and other structures


can be used for this purpose as long as they can be demonstrated to
maintain planarity. Two non-parallel holes yield as many as four
possible solutions, which can often be reduced to a single acceptable
solution from minimal knowledge of the region. If one of the holes
is vertical, its a value gives the true dip and only the strike is in
question. For a unique solution, three holes are required. The
basis for the solution, as discussed by Phillips (1971), is the
Exploration of Rock Conditions 147

NORTH

,H //
\ //
4~o / / /'
, /'

Plane P~ \ //
(required) ',- . __ ---
MIN (P) _ _ __ ____ -~MAJ (P)
-------- \
\
\
\...
,
\

(b)

observation that the true orientation of the normal lies in a double


cone of possible orientations of radius (90 - a) centered about the
drill hole axis and its opposite. The solution is given by the
unique intersection of cones. Since we attach no directionality
to the normal to a plane, for the purposes of this exercise, a
complete mapping on one hemisphere is sufficient. Figure 4 - 28
presents an example. The core-bedding angle was measured in three
non-parallel drill holes as follows:
Good man-GeologICa l Eng -11
148 Exploration of Rock Conditions

NA = N 8 W, 60° NORTH
NS=S8W,60°
0
NC = S 17 W, 50
ND=S75E,60°
0
N E = N SOW, 18
0 Hole 3
N F = N 62W, 11 90 - ex
= 46°

Hole2~
0
90 -a. =45 0

~Hole 3
o 90-a =46
0

Figure 4-28. Absolute orientation of discontinuities from a angles (figure 4-25a)


measured in three non-parallel drill holes.

hole orientation core - bedding core - normal


number angle (a) angle (90-a)
0
I vert ical 60° 30

2 50° to the west 45° 45 0

3 24 0 to the S33E 44° 46 0

Small circles of required radius are constructed about the hole axes
yielding 6 intersections. (In the case of drill hole 3, a part of
the cone about the upward direction of the hole is required as the
Exploration of Rock Condit ions 149

downward axis defines a c o ne e xtendi ng int o the uppe r he misphe r e . )


NB, the only simu l taneous in t ersect ion of a l l th ree con e s, is the
solution, giving the strike of the plan e a s N82W , and dip 30 0 t o t he
North. Had only holes 1 a n d 2, or 2 a nd 3 b een us e d, th e r e wou ld
have been two possible solut ion s , i n t h is i n stance wi t h vast l y
different strikes.
Frequently a few parallel p l ana r fe atu res will recu r in a number
of holes. If the core recovery h as b e en exce llent , so t h a t f e w
reference line int e rru p tions occu r, these fea t ures can be us e d t o
orient the refere nce line, as s h own i n f igure 4-27a , thereby al l owi ng
all the core to be or ie nte d wit h o u t any s pe c i al survey s . If t his
is not feasib le , then core o r i e n t i n g de v ices can be us ed during th e
drilling operat ion.

Core Orientation Devices

In shallow inclined boreho les a t Mt . Is a , Au st ra l ia , Rosengren


recorded the absolute orient a t ion o f core s i mply b y bre ak i ng a
bottle of paint a g a i n st t h e e nd of the h o l e . The paint r a n down
the core st ub , ma r k i ng t he d i p o f t h e S plan e on the e nd of th e c or e .
In holes inclined mo re than about te n degre e s f r om t he ve rt ical , a
simple and effecti v e mechanica l orientati on device -- the Craelius
core orienter (Atlas Copc o ) -- can b e use d in h ard rock, Thi s dev i ce
(figure 4-29a) is a cylinder, o f a bout the same diame t e r as the cor e,
with six locking extension fe e t . At t h e beg in ning of each core run ,
it is inserted on the f r ont of t he core ba rre l with the fe et ful ly
extended. It is lowe red with the core barre l and when it hi t s t h e
bottom, the feet are dep res sed d i ff e re nt ially unt i l they l ock i n to
position, while a fr ee a luminium b a ll is i mpre ss e d agai nst an
aluminium marking d i sc, c a u sing an i n de ntati on def i ning t he true di p
of the S plane. Th e o rie nting dev ice ri de s int o t he barrel as c oring
progresses. When the core b a r rel is empt i ed , t h e t op o f the c ore is
laid in an alignme nt cradle (figure 4- 29b) again s t t he Cr aelius
device and rotated to fin d t h e p r oper fi t of t h e feet a ga i n st the
rough top-of-core. The a ngl e Sa bet wee. n th e d i p of t he S pl ane a nd
the reference line establis h es the or i e ntation of the r efe r enc e line
for the contents of t he c o r e b a rrel . When using t he Crae l ius co r e
150 Exploration of Rock Conditions

- - -- Pressure Rod
Steel Pins

..---- - Pin Retaining


Springs

Pin Cylinder

Ball Chamber
~LOCking Wedge

Spiral Spring

(b)

Figure 4-29. (a) The Craelius core orienter (Atlas CopcoABEM).


(b) Top portion of core aligned with bottom of Craelius core
orienter in an alignment device. Note the indentation in the ring
forming the right end of the orienter; this indicates the dip of the
{3 plane when the core is inclined parallel to the borehole.

(a)

orienter, it would be most convenient to draw the reference line at


the dip of the S plane.
The Christensen-Hugel barrel contains three knives on a shoe
mounted on the end of the inner barrel, so that as the core enters
the core holder, three grooves are cut longitudinally (figure 4-30).
The barrel also has an Eastman, Multishot directional survey instru-
ment, which photographs a compass giving the bearing and plunge of
the hole, and the orientation (the angle S) of a marker oriented
relative to one of the scribing lines. The determination of atti-
tudes of planar features in the core can be made using the method
previously given, or by placing the core in a goniometer (figure
4-30). Voloshin, et al (1968) described application of this equipment
by the California Department of Water Resources for investigation of
stability for surface excavations in shale. This approach is more
expensive than the Craelius device but better suited to the softer
rocks.
Exploration of Rock Conditions 151

The "Integral Sampling Method"

This method, developed b y the Nat ion al Ci vil Eng i n e eri ng


Laboratory of Portugal (LNEC), ( Ro c ha , 1 971), return s a complet e
sample in perfect relative orien t a ti on b y ov erdr i ll in g pre - reinforce d
core. This is accomplished by c e me n ting a r e in fo rc in g bar in a
co-axial hole of smaller diameter pre -d ri ll e d i n t he botto m o f t he
borehole (figure 4-31). The core o r i e n ta ti on i s kn own if the orien-
tation of the positioning rods is me a s u red a t t he t ime t he r e i nfor c -
ing bar is installed. The method is expen sive b e c ause t h e dri l ling
must be interrupted until the ceme n t h a rde ns a nd t he orient a t i on
through the intermediary of positioning r o ds i s p r o b abl y n ot reliabl e
beyond a maximum depth of about 30 0 f e et. However , as a spe cial -
technique addressed to specific q uestio ns , e .g. c oncern i ng the
reason for core loss at a critical lo c a tio n or of t he true n a ture of
a known fault zone, this is a p a r t icu lar ly i mp ortan t t ool. For

Figure 4-30. Orientation of core, grooved


with the Christensen-Huge l barrel , being
reconstructed by means of a core goniometer
(courtesy of Christensen Diam ond
Products Co., Salt Lake Ci ty).
152 Exploration of Rock Conditions

Figure 4-31. (a) The I ntegral Sampling


Method; from Rocha (1971). (b) An
integral sample of decomposed granite;
from Rocha (1971). (c) An integral
sample taken across a gouge-filled fault.
,
uu Ull u
'" . ,
/..

' ~
/
"

/ ,
:

(al

COAXIAL
BOREHOLE

(a)
DOD U
(0) (el (d ) (el

GRAN' ''~I C RES ID UAL, SO IL - 75 em BOREHOLE

(b)

SC H ISTOUS ROCK - 75c'm B O R EHQ L E

(c)
Exploration of Rock Conditions 153

example, Rocha and Barroso (1971) upgraded the evaluation of a dam


foundation in fractured hornfels after examin ing the c on tinuity,
smoothness, and fill in g of fractures in an integral sample. The
integral samples allow measurement of joint apertures; in figure 4-32,
from Rocha (1974), such data are usefully presented o~ a polar equal
area project ion by using different symbols for joint normals in each
of four aperture r an ges. The application of integral sampling to
site investigations will greatly increase our knowledge and apprecia-
tion of discontinuous rock masses.

Geophysical and Optical Orienting Devices


Core orientation is possible with a photographic record of the
borehole wall obtained with a borehole camera. For example, the NX
boreh o le camera manufactured by Republic Engineering and Manufacturing
Co ., St. Pa u l , Minnesota, p h otograp hs th e ima ge d i splayed o n a c o n ic a l
mirror, whi ch l ooks th ro ugh a cylindrical window. The film drive is
timed with a strobe light as the camera is lowered, exposin g frames
every 3/4 in c h as the camera is hoisted up the hole. Interpretation
of the result in g photos for planar orientation was described by

N
o

W 90 t-----t--- -t- t--- f---f 90E

Figure 4 -32. Joint attitudes and


apertures measured in schists at a • < 0.1 mm
Portugese dam site using the 30 CD () 0.1 - 0.3 mm
~0. 3-0.5mm
integral sampling method; repro- o {
duced from Rocha (1974). 0>0.5 mm
5
154 Exploration of Rock Conditions

Trantina and Cluff (1963). Borehole television devices, e.g., the


Eastman F.B. 400*, allow detailed mapping of the borehole walls in
the field. Portability may be more restricted than with a borehole
camera, and surveys consequently may be more expensive. Borehole
periscopes, e.g., Eastman BP34, allow measurement of planar orienta-
tions down to maximum depths of the order of 100 feet. Direct
viewing optical instruments such as this are particularly valuable
for study of discontinuities behind the walls or roof of an under-
ground opening or at shallow depth beneath a foundation . The dis-
advantage of all optical instruments is that they do not work in
holes filled with muddy water. Geophysical loggers with orienting
capability can overcome this difficulty.
A borehole logging probe which permits orientation of planar
features is the Birdwell Seisviewer** described by Myung and Baltosser
(1972). It contains a rotating acoustical transducer, pulsed 2000
times per second. Reflections are received by a detector and trans-
mitted, together with a north marker provided by a built-in magne-
tometer. On each revolution, the north marker triggers a sweep on
the viewing oscillograph, and the sweep is intensity modulated by
the reflected acoustic signal. Advance of the logging tool down the
borehole produces additional horizontal traces separated from each
other in depth. A typical picture, photographed from the visual
display, is presented in figure 4-33; it can be thought of as an
unrolled borehole with the longitudinal cut at magnetic north.
Because damaged rock gives poorer reflections, major fractures
appear as dark curves. Geophysics can also aid in core orientation
through the measurement of magnetic remanence stratigraphy (Zimmer,
1963) and by means of Dipmeter logs. The Dipmeter is a borehole
probe providing continuous resistivity versus depth logs simultaneous-
ly along three radii at 120 0 angles. When the instrument passes an
inclined bed or significant discontinuity of anomalous resistivity,
distinctive patterns repeat on each trace at different depths. The

* Eastman International Co., GMBH, Hannover-Westerfeld, Germany.

** Birdwell Div., Seismograph Service Corp., Tulsa, Oklahoma; the


instrument was previously called the "televiewer".
Exploration of Rock Conditions 155

Figure 4-33. The image of high angle


N E S w N
1
-fractures intersecting a borehole as 5560 - -- I

logged by the Birdwell Seisviewer;


reproduced from Myung and Baltosser
(1972).
5561
0
.. 58 N 70 E

5563 .. 74
0
N 45 W

5 564 -

5565 -

depth separations of an anomaly allow calculation of two apparent


dips which determine the attitude of the plane, as discussed by
Phillips (1971) and DeChambrier (1 95 3).

Other Downhole Measurements

The investigation of dam sites frequently includes pump-in


water pressure tests, from which the rate of steady flow (q) from
a packed off section of the borehole can be plotted against the water
pressure (~p), as discussed by Lugeon (1933), Sabarly (1965), and
Maini (1971). Sabarly suggested programming the test for a series
of steady flow measurements at successively higher pressures and
then at successively lower pressures, as shown in figure 4-34. In
this figure, q is the flow rate (e.g. cubic meters per second) and
~p is the differential pressure -- the difference between the down-
hole pressure inside the borehole (corrected for head losses) and
the original static water pressure in the rock at the test section.
The shapes of the graphs of q(~p) help to identify the phenomena
occurring. Laminar flow gives a response like that of figure 4-34a.
Turbulent flow, possibly explained by an open fracture, or by a leaky
packer, may give a curve resembling figure 4- 34b. Washing out or
opening of a fracture, or rupture of the packer, may give results
156 Exploration of Rock Conditions

q q

~
y

~p ~p
(a) (b)

q q

~p ~p

q q

Figure 4-34. Different types of


response to pump·· in water
pressure tests and their interpreta- tip ~p
tion; after Sabarly (1965). (e) (f)

like those of figure 4-34c. Figure 4-34d can be explained by plugging


of fractures or pores in the rock by fines in the water. It is
important to use clean water for these tests. Figure 4-34e may be
explained by plugging of the flow paths in the rock at low pressures,
followed by opening of fractures or washing out of fillings at higher
pressures. Reversible opening of a fracture as the normal stress is
reached and exceeded by the down-hole pressure may explain figure
4-34f. A proportion of such tests will intersect no conducting
fracture, and consequently will yield no flow at any pressure. Snow
Exploration of Rock Conditions 157

(1968) used the percentage of n o - flo w te st s ac cordin g to th e Poisson


distribution to calculate th e mean spac i ng b e twe en permeable f ractures .
Borehole logg i ng by geoph y si cal tech n ique s c an reve al impo rt ant
discontinuities and permit t h e ir co rre l a tio n from h o le to hol e .
Equipment is now available to u s e st an dard t echn iques of t he p e t rol eum
industry in NX b oreholes. For e xamp l e , t h e Wideo model 1 200 resis-
tivity and spo nt aneous potent ia l l ogg e r is a portable , batte r y
operated u n i t h avi n g a s o nde 1 - 5/ 8 i nches in diameter. By a hand
operated winch , it c an operate to 1200 f oot dept h s . The Birdwe l l
"3-D Veloc it y Logge r" is avai l a bl e with a 2-1/ 4 inch diameter son de
which encompasses a seismic s ource a n d receiver as clos e as one f oo t
apart, (Geyer and Myun g , 1971 ). I t c a n a lso b e used i n a hole t o
hole configurati on . The dev ice p roduce s a re cor d of a mp l itude versus
time of the transmitted ene r gy a t su ccess i ve de p t hs and can be in-
terpreted to yi eld the long itu din a l a nd shea r wave ve locit i es o f
the rock around the b o r eho l e . The relat ive a mp l itude s o f shear and
compressive signals can be used to evalu a t e the mechanical p rop e rt ies
of fractures, espe ci ally if i ndiv i du a l f ra c tures are separately
logged, e.g., by a Seisviewer or o t h e r frac t ure orie nta ti on instru -
ment.
It is also possible to a pp ly s ta tic l oads to t he walls of t he
borehole an d measu r e d irec t ly the de fo rmab i l it y a n d e l ast i c ity of
the wall rock. Some in st r uments , e .g. , t he LNEC dila t ometer, (Rocha,
et al., 1969 ) , apply a radial p r essure i ns ide a rubb e r tube and
measure the radial expansion of the boreho le. Boreho l e jacks) such
as the "Goodman Jack," f or ce steel plat e s to e xpan d along a di a me t er
and monitor the resulting borehole s t r et ch (Go odman , Va n, and He uze ,
19 72) . If the hole is l o gge d to r eve al t h e n u mber of f ractures i n
each test section, with ei the r t ype o f dev ice i t is p ossible to st u dy
the variation of rock mass def o r mabi lit y wi th severity o f f racturin g
in-situ.
The developing technol o gy of e xp l o rat ion makes it possible t o
extract a great deal of inf o rmati on about t h e l oca t ions , orien t at i ons ,
s paci ng s, an d propertie s of di scon t in u ities i n rocks. The rema i nder
of this book is conce r n ed wit h how des cr i ption an d properties o f
discontinuities can be i ntrod uce d i nto e ngi n e ering analysis.
5
mechanical properties
of discontinuities

DETERMINATION OF PROPERTIES

Discontinuous rock masses can be analyzed in either of two ways.


The weakening and softening influence of the network of discontinui-
ties can be accounted for implicitly in calculations by modifying
the strength and deformability properties assumed for a large body
of rock. Or, the actual properties of individual discontinuities
can be introduced explicitly in the analysis as discrete elements of
the rock mass. In order to characterize each joint set, physical
properties may be estimated on the basis of detailed descriptions.
Or, after a program of field sampling, they may be measured in
laboratory experiments. It is sometimes feasible to characterize
joint properties through tests conducted in the field. In special
cases, properties may be "back calculated" from deformations observed
in jointed rocks in the field.

Samples of Joints

There are several ways of obtaining samples of joints for


laboratory testing. Suitable specimens of natural joints might be
found in the core box, and prepared for the shear test by trimming.
Usually, however, special drilling will be needed expressly for
sampling~particular discontinuities. In a rock face, it is some-
times possible to orient a thin walled coring bit parallel to a
discontinuity to provide a longitudinal joint sample (figure 5-la).
Mechanical Properties of Discontinuities 159

It is more difficult to sample by dr i lling perp e nd i cu l arly to a


joint as the core tends to turn on any i nitially open crossing joint
grinding off the asperities. Pre - bol t i ng c an be used for join ts at
shallow depth (figure 5-lb). For tria x i a l t est i ng, as discussed
later, it is desirable to orient join t s at ab o ut 30 degrees t o th e
core axis and this can sometimes be don e s atisfactorily in the fiel d
by orienting the core barrel appropriat ely; a t r ip l e tube barrel is
especially useful for such an effo rt (se e Chap t e r 4).
Another approach to samp l ing of j oint s is t o e xtract a b lock
containing an undisturbed joint ( figu r e 5 - lc ). For soft seams ,
pre-bolting is necessary to retain th e seat ing of th e seam during
extraction of the block. Extremely sensi ti ve seams can be e xtract e d
by wire sawing ("S.E.I.L." method, Ho ek a n d Bray , 1974) .
An entirely different approach is to construc t a n art ificial
joint in the laboratory. One can introduce a r ough or smooth join t
in a sample of the actual rock by s pl it t i ng t he rock in a Brazilian
test (figure 5-le), diamond sawin g ( figure 5-1f ), or producing
a planar shear failure in a triaxi al te s t. Or one can mold part
of a rock outcrop with dental mol ding p las tic (e.g . "Jeltrate t l ) or

plastic or ru bber moulding

e;;;?
compound
3 clay dam
4 rock outcrop
( ol (d)

--
(j)
diamond
saw cut

shallow
groove

Figure 5-1. Types of joint


specimens a, b, c : natural joints d, e.f · artificial joints
160 Mechanical Properties of Discontinuities

~pore
pressure
input

heat treated
50CrNi 13
soft steel shield

ItnriiJImjr-r--++- jacket

- confining
pressure
input

fluid
TIlrITll'It--I----it---+- dis t ri bu t ing
disc

pore
~ pressure
meter
Figure 5-2. Triaxial test chamber;
cm ! !! ! scale (courtesy of Dr. F. Rummel, Ruhr
o 2 3 , 5 University, Bochum, Germany).

"potting compound" (e.g. "Dow Corning Silica Set No. 105"), after
which the mating rough surfaces are cast with hydrostone, sulfur
capping compound ("cylcap") or epoxy (figure 5-1d); this allows
repeated virgin tests with different environmental conditions, for
example different filling material characteristics (Schneider,1974).

Laboratory Triaxial Testing of Joints

The most generally available equipment for jOint testing is a


triaxial testing chamber. Figure 5-2 shows a triaxial cell with
sufficient internal space for shearing of joints. The joint triaxial
test was introduced by Jaeger (1959), and results of triaxial tests
on joints have since been reported by Lane and Heck (1964), Handin
and Stearns (1964), Raleigh and Paterson (1965), Byerlee (1967),
Rosengren (1968), Heuze and Goodman (1967), Goodman and Ohnishi
(1973) and others.
Consider the triaxial specimen with a joint inclined at ~ with
the long axis (figure 5-3a). First, an all around pressure equal to
(a)

T Critical Mohr circle for


Joint slip for:
OJ
1/1 = 1/11 or 1/1 2
+

(b)

Sh ear strength
curve f or j oint

Stress path on
Joint Figure 5-3. (a) Application of Mohr's
1st stage/ /
/2nd stj-lge circle for examination of the tendency
to slip on a discontinuity. (b) Stress
/ /3rd stage path during a multi-stage triaxial test.
/ / (c) Stress versus deformation duri ng
/ the mUlti-stage triax ial test.
(d) Deformations during the mu lti-
stage triaxial test.

(c) (d)
x tJ.z
,... ____ 1 ___ ,
v I I

tJ.x

Unbonded
strain gage

Specimen Shortening
162 Mechanical Properties of Discontinuities

°3 is applied, and then maintained as the deviator stress 1 - 3


) (° °
is raised. Slip occurs when point A, representing the shear and
normal stress on the joint plane first touches the shear strength
curve for the joint. If Wis near 0 degrees, or larger than about
70 degrees, the Mohr circle may become tangent to the Mohr envelope
of the rock before the joint can slip, as discussed by Jaeger (1959),
so that Wmust be controlled within definite limits. The normal
and shear stresses on the joint are, theoretically,

(1)

These equations, (1), can be combined to yield:

T (2)

Thus the stress path during a triaxial test with a joint is as shown
in figure 5-3b. When slip is initiated at A, the confining pressure
is quickly raised from G
3,
°
0 to 3 1 and then the deviator stress is
,
raised provoking slip at B, etc. This test may therefore be termed
a "multistage" triaxial test.
The results of such a test can be useful only if the specimen
is free to slide along the joint without frictional restraint from
the ends. Rosengren (1968) showed that a friction coefficient K at
the ends adds new stress contributions on the joint of and TF , in
the first loading stage, equal to:

K °1 sinljJ cosljJ

and (3)

When added to equation 1, the effect of friction proves unacceptable


unless K < 0.01. This can be achieved using a pair of smooth, flat
plates on each end of the specimen lubricated with molybdenite grease
Mechanical Properties of Discontinuities 163

(Rosengren, 1968) or polished to a fine finish (Wawersik, 1973),


(see figure 5-2).
Vertical and horizontal displacements meas u r ed in the triaxial
test are inclined with respect to local coordinates normal and
parallel to the joint (figure 5-3d). The joint shear displacement
~u and normal displacement ~v are obtained fr om t h e vertical an d
horizontal movements ~z and ~x across the joint:

~z cos~ + ~x sin~

and (4)
~v ~z sin~ - ~x cos~

where ~z eq~als the total specimen shortening mi nus t h at due to


compression of the rock and ~x equals the total s p ecimen l ateral
deformation minus that caused by lateral strain of the rock. ( Only
the volume change of the rock influences the volume of the co n fining
fluid so the rock deformation can easily be separated f r om the total
diameter expansion indicated by a circumferential unbonde d
extensometer.)
A correction is necessary to account f o r cont in uous ch a nge in
the area of contact as a result of shear displacement. The t rue
contact area across the joint, after Rosengren (1968), i s :

D2 (28 - sin 28)


(5)
4 sinlJl

-1 ~u simp
where 8 cos D o < 8 < TT/2

and D is the diameter of the cylindrical specimen. The arg o f cos -1


~ (~x)/D ~ ~z (tan~)/D

When ~ u/ D is small, Rosengren suggested:

TTD2
4 sinlJl - D ( ~u) (5 a)

As shear progresses, the area of the joint which is no longer i n


Goooman- Geological Eng. - 12
164 Mechanical Properties of Discontinuities

rI I ~+RI ~
U t ~ ~ clamp
G) coeficient of I ri:l
friction = K I :~ rubber jacket

- rr-J-'-------"L..-_ _ --r-': rJ
I
I
I---------il
I
I
I
o-c

(~v)
not shown

/ R I
/ 21
I

r-J
I
/
I / PI / RI
,..:.J
I
I
--
I / / / ,-
!r®c1amp
Figure 5-4.Joint triaxial test with large
C I
I
///
/ /
I
I
displacement; after Rosengren (1968). I

contact becomes exposed to a normal pressure equal to the confining


pressure (figure 5-4). The resulting curves of 01 - 03 versus
specimen shortening, z, are consequently difficult to interpret. The
test data should therefore be transformed to display T and ° versus
nu and nv. Consider the equilibrium of the bottom half of a jointed
triaxial specimen under confining pressure °c after shear displacement
nu, as shown in figure 5-4. The normal stress ° and shear stress T
in the joint are given by

0A N
c
and (6 )
TA c T

where
Mechanical Properties of Discontinuities 165

7TD2
P (5
{i
1 c

2
P (J D /tanlJJ
2 c

(~
D2
P
3
(5
c sinlJJ -AJ C:! a D (nu)
c

Rl axial deviatoric force (load read on testing machine


dial minus P ), and
1
friction force on platens K(R + (5 A sinlJJ)
I c c
(A is given by equation 5).
c

Laboratory Direct Shear Testing of Discontinu ities

The direct shear test (figur e 5- 5) is a natura l way to test


properties of discontinuities especially at low normal pressures.

(0) (b)

(c) (d)

Figure 5-5. Controlled normal stress (a and c) and controlled normal displacement (b and d) shearing modes.
166 Mechanical Properties of Discontinuities


T

Figure 5-6. Scheme of the Berkeley direct


shear machine; from Goodman and Ohnishi
(1973). The test is conducted inside a water-
tight chamber (1) by advancing a piston (3)
driving the bottom half of an inner shear
box (2).

A sample is cemented in a shear box, using hydrostone, sulfur capping


compound (cylcap), or epoxy. The surface of discontinuity is
coincident with the plane of shearing and a gap of about one centi-
meter (more if the joint roughness is great) is left unbonded between
the upper and lower boxes. Normal load is applied by a hydraulic
ram, air cushion, or screw and maintained while shearing load is
, built up, either by a screw or a hydraulic ram. The normal load
should move laterally to remain centered over the contact area. Joint
thickening (dilatancy) or thinning (contractancy) can be measured
directly if the normal load is very soft (figure 5-5a). If the normal
load is stiff (figure 5-5b), a tendency for dilatancy will be deter-
mined by monitoring the change in the normal load during shear. Most
machines in use prohibit any specimen rotation during shear. Specific
direct shear machines are described by Krsmanovic and Langof (1964),
Lombardi and Del Vesco (1966), Evdokimov and Sapegin (1967), Locher
(1968), Bernaix (1969), Hoek (1970), Rengers (1970), Burman (1971),
and Goodman and Ohnishi (1973), (figure 5-6). An economical portable
direct shear device developed at Imperial College, (Hoek and Bray,
1974) is available commercially.* Objections to the direct shear
test on intact rock specimens arise from the unknown local stress
conditions and failure mode by en-echelon_tension cracks. However,

* Robertson Research, Llandudno, North Wales; and Golder & Brawner


Assoc., Vancouver, Canada.
Mechanical Properties of Discontinuities 167

a finite element study of stress conditions in a shear box by Kutter


(1971) showed that a uniform stress distribution exists over most of
the joint surface in the case of tests on seams or preexisting
joints.
The direct shear test yields directly the relationships between
stresses T and 0 and displacements ~u and ~v necessary to character-
ize the deformability and strength properties. A typical test record
for a sample with a joint is presented in figure 5-7. The parameters
to be extracted from such tests will be considered later.
It is usual to program a direct shear test so that the normal
pressure remains constant during shear as in figure 5-5a; this
corresponds to sliding of a free block on a slope (figure 5-5c).
However, such a test yields too Iowa shear strength for sliding of
blocks constrained between parallel dilatant joints, as for example
in an underground opening (figure 5-5d) or in the middle of a rock
wall. A no-displacement or controlled stiffness condition in the
normal load direction, which can be obtained by servo or manual
feedback control, is the correct type of shear test for such design
situations. The results of a normal displacement controlled shear
test will depend not only on the stiffness of the normal load member,
but upon the stiffness of the wall rock so the data require pro-

Normal Stress, (/n (psi)


200 400 600 800 ~1600
tf)
a.
c
E~ 10
QJ CI1
f-> 1200
tf)
U QJ CI1
o.c QJ

~
- U
~= 20 800
0'"
-0
0 ......
E 30 ~
o
Z
40
o 0.10 0.20 0.30
Shear Displacement, U (inches)
Figure 5·7. Direct shear results
for an intact specimen (1) and a
c:
specimen with an initially open ~
E-;;:; 10 ~

discontinuity oriented in the plane QJ (lJ ~ 20


U.c (lJ
o u
of shear (2). First the normal force ~= 20
ct 0 f---L..---,,......,,.--t-~~..&.--~~...
QJ 0 .20 0.30
was applied (upper left); thereafter
· _rfl
01
, -0
&-20 Shear Displacement,u
0"-< (inches)
the specimen was sheared at con- E ~30 ~ -40
stant normal pressure. From Z
o ~ -60
Goodman and Ohnishi (1973) . ......
40
168 Mechanical Properties of Discontinuities

cessing to scale to any given field case. One of the advantages of


the direct shear test over the triaxial test is the greater ease
in specifying and controlling the normal load boundary condition.

Field Shear Tests

The direct shear test can be conducted in situ (figure 5-8).


The procedure involves selective excavation to isolate a test block
on a plane of weakness, either in a gallery or on the surface. In
the former case the walls and roof of the chamber provide reactions
for the normal and shear forces, (see e.g. Serafim and Lopes, 1961).
On the surface, a cable anchored beneath the center of the block,
(Zienkiewicz, 1966; Haverland and Slebir, 1972) provides reaction

Figure 5-8. I n-situ direct shear test


at Auburn dam site; from Haverland
and Siebir (1972).
Mechanical Properties of D iscon tinuities 169

for the normal force or the test is con du c t ed with the b lo c k' s self
weight alone (Ruiz and Camargo , 1966 ). Shear f o rces c an be arran ge d
by jacks across a trench behind the rear face of the test bl ock. In
soft rocks such as coal , marl and soft s h a les, it may be pos s ib le to
jack into place a shear box equipp ed with cutte rs (Brawn e r et aI ,
1972) ; this provide s s up p ort for the shear bloc k duri ng i t s pre para-
tion ove rcomi ng the d amaging eff e cts of t ota l dec omp r es s ion.
Haverland and Slebir applied the n orma l force bef ore f r ee in g the side s
of the test block to avoi d total de c o mpres s ion. In situ shear tes ts
are quite expensive; because of scatter in resu l ts a few i n s i t u t e s t
results may have questionable worth unle ss supported by a l abo r atory
test program.

Study of Case Histories

No laboratory or field test, however careful l y contr ol led , c an


duplicate the scale and cha ra cter of the loading, b o unda ry and
environmental conditions inheren t to en gin e ering service. One
hopes through shear tests to eva l u ate compon e nts of t he j oint pr o -
perties, which can lat e r be integrat e d in an a na lys is or mode l study .
But a relevant case his tory can he lp to ins ure t hat t h e extr ap olation
from test to fie l d is basically correct. The e x tr act i on of prop er-
ties of joints f rom " back calculations " p rese n ts a prob le m i n ver se
to that of design an d in general can n o t yield un i q ue a n swe rs.
Therefore the most useful case hist o r ies are those t h at re s e mbl e t h e
wo rk in question. For limiting eq u ili b ri um probl e ms, na t ura l an d
artificial rock slopes can provide con side rab le help in ev al u at i ng
properties of joints. In this connection Hoek a nd Br ay ( 19 74 )
developed a series of charts and functions greatly f a ci li t a t ing back
calculations of slope failures. As di s c u s sed later, it i s poss ible
to approximate the relation between r e s i d ual she a r st rength and
normal stress for joints sufficiently well for some ap p lic ati ons by
the Coulomb equation:

S ( 7)

The residual fricti on angle ¢ c an be es tab lis hed b y s h e ar t est s and


r
170 Mechanical Properties of Discontinuities

field observations, but the cohesion, C can not. Therefore, deter-


J
mination of reasonable ranges for probable values of C may be the
J
main object of case studies. On the basis of more than 40 case
histories, Hoek and Bray were able to suggest a range of values for
the cohesion of rock masses (Table 5-1); these values help to place
in context cohesion values obtained for individual weakness planes.
Finite element analysis, model studies, and other analytical
methods as well as graphical solutions are used in calculating field
case histories. These methods are to be discussed in later chapters.

DEFORMATIONS IN JOINTS

Normal Deformations

When a block is placed lightly on a rough surface, the propor-


tion of tne surface area in actual contact is almost zero. The
entire contact force is sustained at three or more point contacts.
Under increasing normal load, the point contacts enlarge by elastic
deformation, crushing, and tension cracking, while the deformation

TABLE 5-1
Order of Magnitude of Joint Cohesion for Rock Masses
(after Hoek & Bray (1974) figure 70)

2
C (psi) C (Kg/cm )
J J
Soil < 56 < 4

Weathered soft rock;


Discontinuities in
hard rock 56 - 140 4 - 10

Soft rock masses or


jointed hard rock disturbed
by blasting or excess loading 140 - 280 10 - 20

Undisturbed jointed soft


rock masses 280 - 420 20 - 30

Undisturbed hard rock


masses 420 30
Mechanical Properties of Discontinuities 17 1

brings new regions into contact. It is poss ib le to pursue t he se


mechanisms mathematically and develop a theory of normal deformation
under increasing normal load, as was done for metals by Bowden and
Tabor (1964); however the system is so poorly defined that an
empirical approach is more useful.
There are two physical constraints on normal deformations in
discontinuities. First, an open joint has no tensile strengt h.
Secondly, there is a limit to the amount of compression possible,
a maximum possible closure, V ,which must be less than the "th ick-
mc
ness" of the joint, e (figure 5 -9 a). But presently we can only guess
the relationship between Vmc and e. Combining these two conditio ns
demands that we fit the normal pressure-deformation relatio nship into
a quarter space as shown in figure 5 - 9b. A simple relationship
satisfying these conditions is the hyperbola :

o - t;
A ~v ) t (~v < Vmc ) (8)
t; ( Vmc ~v

where t; is the seating pressure, defining the initial condition for


measuring the normal deformation ~v. The continuous curvature of
o = f(~v) described by (8) can be observed experimentally as shown
in figure 5-9c. Curve A, in the left half of the figure, shows
the shortening of an intact cylindrical specimen (3 .6 inches long
by 1.75 inches in diameter) on its third cycle of loadin g . The first
load cycle showed large hysteresis and inelasti city , b ut the second
and third cycles of loading of the intact specimen produced almost
identical, elastic compression curves . Then the specimen was turned
on its side and compressed between knife edges creati ng a single,
rough and wavy extension fracture parallel to its en ds. The specimen
was reassembled and recompressed yielding curve B (figure 5-9c) .
The difference between the compression curves for the jOinted and
intact specimens descr i bes the compression of the mated jo int; it
is plotted in the right half of the figure.
At t he en d of these loadin g cycles, th e wall ro ck a nd join t
showed no visible damage. Then, the up per block was rotated to
create a mismatched joint, with point contacts and mean apertur e (e)
172 Mechanical Properties of Discontinuities

Figure 5-9. (a) Idealization of a joint.


"- '-.. "- /' / I
(b) Behavior of a joint in compression.
\.. \...
~
./ I /'

,
"- '-
~ is the seating load. (c) Normal com-
pression of an extension fracture in a
granodiorite specimen. "-
--
/ "- ../
~

"-
/'
-
I- '-
e

(a)

0- compression

b.v

(c)

5000.-----~--·----r-----~------r---~ .-----~------~-----r------~--~

4000

.iii
-9:3000

E 2000
o
z

1000

o 0 .005 0010 0.015 0.020 o 0.005 0.010 0.015 0.020


Axial Displacement (in.) Jomt Closure (in .)

approximately 0.05 inches. Compression of this sample produced


curve C; at point P on this curve, the rock began to split lengthwise.
The test was discontinued; however the curve was extrapolated as
Mechanical Properties of Discontinuities 173

shown. The difference curve, (C)-(A) describes the compression


behavior for the non-mating joint. At the end of the test, about
ten percent of the area of the joint showed the results of rock
crushing.
The following equations (with 6v in inches and a in psi) fit
the joint compression curves (B)-(A) and (C)-(A) of figure 5-9c.
For curve (B)-(A) representing deformation of the mating joint,

V 0.0047 inches, and


mc
6v -. 0004 + .0007 In 0

For curve (C)-(A) representing deformation of the non-mating joint,

V 0.0152 inches, and


mc
6V -.0094 + .0031 In 0

The dimensionless form of the joint compression curve, equation (8),


(with 0=6 7 psi arbitrarily established as the seating load ~) can
be fitted approximately to the data of figure 5-9c. For the mating
joint: A = 3.00 and t = 0.605; for the non-mating joint A = 5.95
and t = 0.609.
The unloading cycles for the jointed specimens followed
essentially the same path as for the intact rock signifying that the
elastic portion of the normal deformation in a rock with a joint is
entirely derived from the rock. Joint compression is essentially
unrecoverable. The highly non-linear, and inelastic deformation of
a discontinuity un de r compression accounts for similar non-linear
and inelastic stress dependency for all properties which are linked
with joint aperture, e.g. electrical resistivity a n d fluid permeabil-
ity. The behavior of discontinuities in compression is discussed
further in Chapter 8 in association with finite element analysis.

Shear Deformations

Krsmanovic and Langof (1964), Hoek and Pentz (1968), Rosengren


(1968), Goodman (1970) and Coulson (1972) discussed the shear stress
- shear deformation curves for discontinuities undergoing shear at
constant normal stress. Rough clean (i.e. "unfilled") joints (curve
174 Mechanical Properties of Discontinuities

(a)

6u

(b)
t

Figure 5-10. (a) Types of


shear stress versus
deformation curves at
constant confining
pressure. (b) Parameters
of shear deformation at
constant confining
pressure. 6u

A of figure 5-l0a) show relatively rapid rise in shear stress to a


peak followed by an irregular post peak history with considerable
loss in load carrying capacity. The ratio, B, of residual to peak
shear stress increases with normal pressure; it may be as low as
0.3 in previously healed or incipient joints at low normal pressure
and 0.6 in open rough joints. Seams or "filled" joints (curve B
of figure 5-l0a) show a convex downward stress-deformation curve
like that of clay, with the peak stress poorly defined and the slope
continuously changing. In a sense, the type B curve is a product of
a history of deformations, which may have started with type A. When
Mechanical Properties of Discontinuities 175

the filling material becomes dessicated, the type B curve is re-


placed by a type A curve, but wetting reverses this response. Most
in situ shear tests reviewed by Goodman (1970) proved to have type
B curves, perhaps because one tends to authorize the great expense
of in situ shear testing only for major seams.
The shear deformation versus shear stress curve for a test
conducted under constant normal stress can be characterized by
elastic, peak, and plastic re ~ ions as depicted in figure 5-10b. The
peak shear stress (Tp) is termed the shear strength while the minimum
post-peak shear stress (T ) is the residual strength. The peak
r
strength demonstrates a scale effect, while the residual strength
may not (Bernaix, 1974). The slope characterizing the elastic region
is termed the unit shear stiffness k (Goodman, Taylor, and Brekke,
s
1968). Joints with type B shear curves generally have lower stiffness
than those having type A curves. Representative values of shear
stiffness for different classes of discontinuities are not generally
known and values are often assumed in analysis. The reason is that
shear stiffness measurements are sensitive to the testing technique
and apparatus, particularly the technique of gripping the specimen
and the location of the displacement measuring instruments. Moreover,
shear stiffness displays a strong scale effect as revealed by Barton
(1972) (figure 5-11).
All of the parameters of joint shear behavior are greatly
influenced by changes in normal stress. The variation of peak shear
displacement and shear stiffness with changing 0 can be simplified
using a model having constant stiffness as in figure 5-12a, or a
model having constant peak displacement as in figure 5-12b (John,
1970). Jaeger (1971) gave examples of direct shear test results
with polished saw cuts in trachyte in which the first loadings fit
the constant stiffness model; however reloading the worn surfaces
more nearly matched the constant peak displacement model. The
variation of peak shear strength with normal stress is described by
the shear strength curve, as discussed later. As shown in figure
5-22, the T -0
p
relationship for a rough discontinuity tends to be
more highly curved than this relationship for the intact rock. For
a restricted range of normal stresses, linear approximations to the
176 Mechanical Properties of Discontinuities

fj. Laboratory

o In Situ
o Model

c:
"-
If)
0..

'-
o
OJ
L
(/)

Figure 5-11. Joint shear stiffness


as a function of the square root of
the loaded area (L) and the normal
stress (25, 100, 500, and 1000 psi
lines are given); from Barton (1972). Test dimension, L (in.)

peak and residual shear strength curves can be used with acceptable
precision. In some cases, the ratio of residual to peak strength
B approaches unity as a increases.

Dilatancy of Continuous Rough Surfaces

A shear test conducted under restricted normal displacement


conditions, (figure 5-13) curves B, will generally yield a considerably
higher shear strength than one conducted under constant normal stress
(figure 5-13) curves A. The reason for these strength differences
is connected with dilatancy. Perfectly mating rough blocks can be
forced to slide past one another only if they are free to move apart,
Mechanical Proper ties of Discontinuities 177

( 1) highest (J
! 1) hi ghest (j

(21 lowest (J !2 1lowest (j

flu flu
(a) constant stiffness model (b) constant peak displa cemen t model

Figu re 5-12. Shear defor mati on models.

v(T)

B
6u(1 ) 6u (t )

(a) Dilation (bl Normal stress

Figure 5-13. Effect of test mode o n


shear deformation curves for d ilatan t
~u(tl joints. A: shear at constant normal
stress; B: shear with condit io n of no
(c I Shear st r ess normal disp lacement.
178 Mechanical Properties of Discontinuities

(to "dilate") to work around asperities; if the blocks are confined,


shearing is possible only if the asperities themselves break. Since
dilatancy can very considerably strengthen a joint, it is an im-
portant property.
This mechanism of dilatancy, involving the over-riding of
asperities, originates mainly from surface roughness. (Another
mechanism -- rotation -- will be considered later.) R~ngers (1970)
measured roughness angles of natural joint surfaces over a band of
steps from 0.01 to 1000 cm, using successively a variable focus
microscope, a profilometer, and terrestrial photogrammetry. Corre-

(a)

L L L L L S= 1· L
ref--~----~----~----~----~--~----~----~

S=2 ·L

Figure 5-14. (a) Example of roughness


angle calculation from digital roughness
amplitude measurements; from Rengers
(1971). (b) Envelope of roughness angles
for sliding to the left and to the right in
the example of fig~re 5-14a; from Rengers
S=4'L
(1971 ).
Mechanical Properties of Discontinuities 179

sponding to each selected step size there is a distribution of


roughness angles; for example, in figure 5-14a, corresponding to
step size S = I·L the surface presents angles u l A through u l G'
o "
varying from +35 0 to -45. Similarly, step S = 2L produces angles
u ,A through u 2 , F varying over a smaller range and so on for S = 3L
2
and S = 4L. Rengers plotted these angles corresponding to the value
of S (figure 5-14b) and constructed envelope curves (solid lines),
assuming that the steepest surface angle of contacting mating blocks
always regulates di latancy during shear with over-riding of asperi-
ties. The envelope of positive angles governs right lateral shearing
(clockwise sens e) while the envelope of negative angles governs
left lateral shearing (counter clockwise sense). The ac t u a l dilatan-
cy during she a ring reflects the cumulative effect of movements along
the rough surface. Since the effective roughness a ng le varies
inversely with t h e di s tance of relative shear motion, the rate of
di latancy decreases progressively as shearing continues. In fact,
t h e dilatancy curve ~V(T) = f(~u(T» is essentially the mirror image

(b)

Dil
~ Roughness angle envelope
+tana
OJ lata ncy curve
3
,,
2 1.0

- tan a
Dil
41 ~

Gooo man-Geo logica l Eng. - 13


180 Mechanical Properties of Discontinuities

of the roughness angle envelope (figure 5-l4b). Barton (1971) made


a similar analysis for rough joints in a model material.
In the field the reference state from which dilatancy begins
to be measured may reflect past shear displacement. In the case
considered (figure 5-l4b) a previous shear displacement equal in
magnitude to L implies a remaining dilatancy of about one half that
inferrable from the analysis of roughness.
Oversliding of asperities without rock breakage is unlikely,
except at zero normal stress; when there is no normal stress or
restraint, asperities presenting angles less than 90 - ¢ can be
~
over-ridden, where ¢ is the friction angle for sliding of flat
~
surfaces of the rock in question. But when the normal stress is
high, the work required to dilate against normal force will exceed
the work sufficient to shear through some asperities. Thus dilatancy
decreases with normal stress and tends to become completely sup-
pressed when the average normal stress reaches the order of magnitude

0.25r-----~.-----~rr--------,-------,-------,--------,
00

0.20~-----+~------~-------r-------+------~------~

::l
0-
"'-.

~ 0.15~------~------+-~~--~-------+------~------~

if)
if)

~
c
.9
if)
c
Q)
0.10
E
0

0.05 ~------~--~~+-----~-r~~---+------~------~

Figure 5-15. Peak dilatancy


angle (i) as a function of the
ratio of normal stress to com-
pressive strength for model o
extension joints; from Barton 5 30
(1971). Peak dilation angle
Mechanical Properties of Discontinuities 181

of the u nconf i n e d compress ive streng t h of t he asp e rities (Bartoo,


19 71) . The she ar strength of an i ndividual asp e rity varie s with i ts
width. Cor re sponding t o e ach value of n ormal load, all asperiti e s
up to a give n base wi dth s uf fer the possibility of rupturing.
Accordi ngly, t he variat i on of dilatancy with i ncreasing normal stre s s
is gove rn ed by a f unc t i o n similar t o t he reduction of dilatancy wi th
in crea s ing s t ep s iz e, (f igure 5-15).
Fiel d measurement of r o ughness , as a function of base distanc e,
can be made phot ogrammetrical l y (Ross-Brown et aI, 1973; Patton,
1966), and by profilomet er (Fecker , 1 970). I f repeated measureme nts
of th e attit ude of an exposed sur f ace are plotted on a stereographic
p roj ection, F e cke r and Rengers (1971 ) showe d that the roughnes s
ang l e s c an be es tima t e d f rom t h e extent of the scatter of poles.
Wh en a ri gid p la t e is l ai d on t he surface of discontinuity , its
a ttit u de will depend upon the relative locations of i ts contacts with
the surfa ce ; thus r e peated measurements taken wi t h a field compass
att ached to a r ig id pl ate will give a range of readings. The rough-
ness an gle , c o rr e spondi n g to a step size equal to the diameter of
t h e p l a te , is e stimated b y measuring the angle be t ween the mean
orientation of t he po le s and t he extreme orientation of t he envelope
t o t he se t of pol e s. F i gure 5-16a gives two examples as well as
comp a risons wi th profilograph data f or two spe c ifi c directions o f
sl iding .
Us e o f a joint pole scatte r diagram for analysis o f roughn ess
is ac cura te only for isot rop i c roughness . When roughness i n a
discon t inui ty is cre ated by planar asperi ties, rather than by
detache d hil ls , t he roughn e ss angles are the apparent dips of the
plana r a sperit ies in the direct i on of sliding.* To illustrate,
(fig ure 5- 16b) con si der s hea r on a surface having roughness formed
0
by pl ane s I and 2 str i king parallel to each o t her and dipping 30
0
and 4 0 in op posite dir e c tions. Th e "envelope" of norma l s is t he
line Nl N2 whe reas the t rue roughness angle e nvelope is the boundary
of the r u led area, cons t ru cte d by finding poles to apparent dips in

* J ohn Boyd , Imperi a l Col l ege, London ; personal communication.


(a)
tan a (±)a
N 1.0
40°

30°
0.5
20°
10°
S= n·L
E 500(mm)
20 50 100
10°
20°
0 o~#I"
'7
...t:j---
Q...~

0.5 ,1'/
30° /
/
/
40° /0
--diameter 5.5em S 1.0
-·-·-dlameter 11 em
............ diameter 21 em tan a 8a
----- diameter 42 em

(b)

LOWER HEMISPHERE
STEREOGRAPHIC PROJECTION
~1 and "'2are normal to asperity planes;
M is normal to the mean plane of the
joint. The ruled area is the roughness
figure based on apparent dips. Line N 1 '
N2 is the envelope to normals, which
has no validity here.

Figure 5-16. (a) Comparison of roughness estimate from the scatter of normals to a single rough
discontinuity Oeft figure) and actual roughness as measured by surface profiles (right figure). The
four scatter diagrams on the stereonet are envelopes to repeated measurements of attitude of a
rough joint with the compass placed on a plate of 5.5, 11, 21, or 42 cm. diameter. After Fecker
and Rengers (1971). (b) Roughness of a discontinuity formed of two sets of component planes
when profiled in different directions.
Mechanical Properties of Discontinuities 183

all vertical planes. For example, for slidi n g parallel to D, the


roughness angle is aD as shown.

PEAK SHEAR STRENGTH

The "shear strength" of a discontinuity refe r s to the peak lo ad


in a test with constant normal stress (which as previously noted may
be quite conservative for dilatant joints under restricted normal
deformation). Shear strength is sometimes called "friction" although
as discussed by Nascimento and Teixeira (1971), no t only surface
friction, but wedging, rotation, and even rolling effects contribute
normal stress dependent shear ~esistance, while cementation and
interlocking can develop additional shear strength.

Surface Friction of Minerals and Rocks

The surface friction of smooth rock and mineral surfaces de ri ves


from micro-interlocking and adhesion, which may require ro c k bre aka ge
for sliding, as well as "ploughing" of harder minerals into a softer
matrix. Jaeger (1971) reviewed motivations, methods, and data of
friction measurements between rock s ur faces. Friction measurements
on smooth rock surfaces were reported by Jaeger (1959), Byerlee
(1967), Jaeger and Rosengren (1969), Coulson (1972), and others;
friction of individual minerals was measured by Horn and Deere (1962).
Friction experiments of even relatively smooth rock surfaces, e . g.
lapped with #400 grit (roughness ~O.OOI in) usually show cons i derable
scatter and are generally more sensitive to changes in moisture
conditions and roughness than to changes in miner a logy. The n otable
exception is the family of sheet silicate minerals, mi ca, chlorite,
clays, talc and serpentine, which exhibit low f r iction, particularly
when wet. Generally the coefficient of friction (tan ¢ = T ja)
fl p
varies in the range 0.4 to 0.8, but in the sheet mi nerals it can be
as low as 0.2 (¢ = 12 0 ) and rocks composed largely of such minerals
l-1
can have quite low friction angles. For example Richa rds (1 9 73)
0
reported ¢ = 20 (tan¢ = 0.36) for moist, smooth surf aces of slate.
fl fl
Drying incre a ses the friction of sheet silicate miner a ls b ut , odd l y,
oven drying significantly lowered the friction of quartz, calcite
and feldspar in Horn and Deere's tests at low confin ing press ure.
184 Mechanical Properties of Discontinuities

Most rock surfaces however are stronger when dry than when wet.
Most specimens tested by Coulson, which included granite, basalt,
gneiss, sandstone, siltstone, limestone, and dolomite, showed higher
friction after displacement of one to three centimeters, accompanied
by secondary fracture of wall rock ("Riedel shears") and formation
of gouge, figure 5-17, (see Lajtai, 1969b), particularly at normal
pressures above 500 psi. Wear of rock surfaces accompanying continued
shear displacement eventually causes the surface to be coated with
crushed material, and it becomes a new kind of specimen. In the case
of dry, unweathered rock surfaces, the new material can bring higher
friction than polished surfaces, but in moist weathered rock surfaces,
wear produces a clay film with a considerable drop in friction; for
0
example Richards (1973) found residual friction of 15 in slightly
0
weathered diorite having peak friction near 30 .
In summary, the friction properties of smooth rock surfaces vary
with micro-roughness, normal load, weathering, environmental condi-
tions, test apparatus, and testing procedure. A value for the
coefficient of friction of 0.5 to 0.6 is a reasonable assumption in
general, but values as low as 0.2 can be expected for rocks rich in
mica or other platy minerals or whose discontinuities are weathered.

Chatter marks
Figure 5-17. Characteristics of a
Compacted debris
gouge zone, artifically produced by
shearing at 1177 psi normal pressure
along sawed joints in Solenhofen
limestone; from Coulson (1972).


Mechanical Properties of D iscontinuities 185

- - ---'------"----- - --N
(a) (b)

Figure 5 -18. Slid ing on incl ined asperities. (c)

Additional Resistance From Sliding on Incli ned Wedges

We have seen previously that dilatancy can cause an increase in


a and thereby stren g then a joint. Sliding asperities can in~rease

the fric t ional re sis tance of rock surfaces even when the dilatancy
does not increase t h e normal stress, as discussed by Patton (1 966).
Cons ider an i d eal wedge-shaped asperity ( fi gure 5-18a) inclined i
degrees above the direction of sliding. The friction angle on the
s l idi n g surface i ts elf is ¢ , i.e.
j.l

T' N' t an ¢
)J
(9)

At the limit of sl i di n g, T and N are connected by:

T T' cos i + N' sin i


N N' cos i - T' sin i
186 Mechanical Properties of Discontinuities

Figure 5-19. Shear VIti


behavior of bidilatant
joints without rotation
(a) dilatancy
under constant normal
stress.

-
~u(tl

~u(l)

(b) shear

introducing 9 and letting tan ¢ TjN gives:

tan ¢ tan (¢lJ + i) (11)

The effect of regular asperities at a uniform angle i is therefore


to increase the friction angle by i (figure 5-18b). This result is
also apparent upon examining the inclination (¢ + i) of the resultant
force on the plane of sliding (figure 5-18c).
The wedge effect is associated with dilatancy,

!:.v (T) !:.U (T) tan i (12)

The inclination i is signed; that is, for left lateral shearing in


figure 5-18, i is negative and (¢) = (¢ - Iii) as the joint shears
with contractancy. Contraction rarely occurs in practice since
actual rough surfaces possess a distribution of both positive and
negative angles causing dilatant behavior when sheared in either
direction (figure 5-19). Only previously loosened surfaces or soft
Mechanical Properties of Discontinu ities 787

seams will exhibit contractancy; in the former case, once the surfaces
displace they will lock and the friction angle will be restored.
If limiting deformations govern de s ign , contractant behavior may
need to be considered. Iida and Kobayashi (1974), for example,
discuss both dilatancy and contractancy effects for computations
of stresses in dams on rock foundations.

Peak Strength as a Function of Normal Pressu re

Equation (11) can not hold at high normal st ress because the work
required to shear through asperities is less than the work to over-
r i de them. Patton (1966) found t h a t a bilinear relation (figure 5-20)
described his experimental data for shear of model joints with
regular teeth. At normal stresses less than aT

T a tan ( ¢ + i) (12a)
P ~

while at normal pressures greater than aT

(12b)

where aT CJ/(tan (¢ + i) - tan ¢ ),


~ r

Figure 5-20. Patton's bilinear cri t erion for shear


strength of joints.
188 Mechanical Properties of Discontinuities

C is the shear strength intercept ("cohesion") derived from the


J
asperities and ¢ is the residual angle of internal friction of the
r
rock comprising the asperities.
Actual rock surfaces obviously cannot be fit by such a simple
model. Jaeger (1971) considered a continuously variable empirical
shear strength equation which can be written:

T (13)
P

When a = 0 the strength curve rises at

while when a is large,

log(p)

Figure 5-21. Jaeger's empirical


shear strength relationship.
Mechanical Properties of Discontinuities 189

Figure 5-2 1 shows h ow su b a f ormul a smooths Patt o n's shear strengt h


criterion ( 12) an d sugges ts a s i mpl e graphical way of ev aluating t h e
paramete rs . Sk e tch a l i n e asympt ot ic t o the peak strength data
plotted in T- a c o ordi na t e s; t h is de fines CJ and ¢ r' and a value of
p f or eac h t es t: p = C
J
+ a t an ¢r - Tp ; t hen -b i s the slope o f
the li ne of l og (p) plott e d ag ains t o. Jaeger found , for example ,
for res i dual sl i ding of n a t ura l joints in andesit e (in psi units):

Tp 270 ( 1 - exp (-0.0015 0 » + 0.41 0

La danyi an d Ar chambau lt' s Equation. While J aeger's empirica l


equat io n shou ld be sati s factor y for a wide range of conditions , it
is help ful to h ave an e q u at ion deri ved f rom i dent ifi able properties
of the joint and wal l rock. La dany i and Archambau lt (1970) combin e d
t he fr ict ion, di l atancy an d i nt e rlock contribut ions t o peak shear
strength to der i ve a g eneral strength e q uation f or discont i nui ties
wh ich h a s pro ved a ccurat e i n mo de l studies. The peak strength is
give n b y:

0 ( 1 - a s) (v + t a n ¢w ) + as sR
T ( 14 )*
P 1 - (1 - a )
s
v t an ¢
W

wh ere as' v , and s R are the fo llowing f unct i ons of a:


a is the prop o rt i on o f joint area sheared th rough t he asper i ties
s
v is t h e dil at i on ra te a t the peak shear stress (secant dila t ancy
rate) ~ v ( T p) /~U (T p) ' an d
sR is t he shear stren gth o f the rock composing the asperities.
Eq u a ti on (14) reduce s to (12a) at l ow a when a
s
= 0 and = t an i . v
P utt ing sR = CJ + a tan ¢ r' equation ( 14) reduces to ( 1 2b) at very
high a where al l t he t e eth a re sheared off , a s = 1, and = O. v
However Ladany i suggested substituting Fairhurs t's paraboli c crite rion
f o r sR :

* The de r i vation follows t he work o f Row e ( 1 962) and Rowe et a l


( 1 96 4). The eq uat ion given in (14) d iffers from t hat presented
by Ladany i i n rep l acing Rowe 's ¢f by ¢w'
190 Mechanical Properties of Discontinuities

+ n - 1 .l.
(1 + n o/q )2 (15)
n u

where q u = the unconfined compressive strength and n = the ratio of


compressive to tensile strength, of the rock comprising the as-
perities.
Ladanyi and Archambault suggested power laws for v and as;
for 0 < _oT'

a (16a)
s

and

v tan i (16b)
o

The suggested values of the exponents are Kl= 1.5 and K2 = 4


(Ladanyi and Archambault, 1972); a increases from 0 at 0 = 0, to
s
1 at a = aT while v decreases from tan io when 0 = 0, to 0 at
a = 0T (see figure 8-17). The transition pressure aT is the normal
stress at which the joints cease to be weaker than the rock itself
and in the absence of sufficient data it can be approximated by
aT = quo ·With these conditions, equation 14 will define a curved
peak stress criterion as shown in figure 5-22.
Barton's Empirical Shear Strength Equation. Barton (1974a)
offered an empirical shear strength criterion for unfilled discon-
tinuities accounting for the variation of dilatancy with normal
stress and the shear strength of the asperities.

( 17)

The factor R expresses the influence of roughness, varying linearly


from 0 to 20 over the range from perfectly smooth to very rough. In
both expressions (14) and (17), the normal stress is the effective
stress if the discontinuity contains a fluid under pressure p; i.e.
on = 0total - p. The unconfined compressive strength qu refers to
Mechanical Properties of Discontinuities 191

the rock forming the asperities. Since weathering is often con-


siderably more advanced along joints than thro u gh the b ody of rock,
q may be considerably lower than values for unweathered rock and
u
it should be obtained from results of tests on the wall rock, e.g.
Schmidt hammer or scratch hardness tests. A comparison of equations
(14) and (17) for rough joints (i
o
= 50, R 20) showed Ladanyi's
pe~k strength to be higher than Barton's in the region 0.5 < Gn/qu <
0.7, and lower elsewhere.

Filled Discontinuities

When soil material completely covers the walls of a discontinuity,


burying all the asperities, shear entirely within the f il ling material
is possible. If the thickness of filling, e, is less than the maximum
asperity height, rock contact will occur after a displacement ~u c

approximated by (Barton, 1974),

~u e/tan i (18)
c

and thereafter the discontinuity will stiffen and strengthen. In


filled, smooth discontinuities, this will not happen an d , in fact,
the peak strength may be less than that of the filling material when
sheared alone at similar rate and confinement (Eurenius and Fagerstrom ,

Figure 5-22. Ladanyi and


Archambault's shear strength
relation for rough joints.
792 Mechanical Properties of Discontinuities

Figure 5-23. Shearing though


a discontinuous rock mass on
compound surfaces.
11 J mean plane

[oj

(1) mean plane

( bJ

1 (21 right lateral sliding,


both cases

~u-
(3J left lateral
,--~-~[L.J left lateral shear case b;
rotational sliding [see textJ

[c 1

1969). This is because the shearing occurs at the rock wall where
pore pressures can not be dissipated and where a small displacement
creates sufficient strain to reach the residual shear strength. Since
the shear strength of filled discontinuities varies over a range from
0 0
as little as 7 , for montmorillonite fillings, to as much as 50 , for
crushed rock or preconsolidated, dry gouge, shear tests are warranted
when the strength needs to be known accurately.

Sliding on Compound Surfaces-Rotational Friction

The peak strength models discussed can describe sliding on a


compound surface made up of sections of two different planes as
Mechanical Properties of D iscon tinuities 193

di sc ussed by Bray (1967), Jaeger ( 1971), Deere et al (1967), and


Ladanyi an d Archambault (1972). If t h e mean plane through su ch a
surface is constructed, the actual surfaces along which sliding occurs
can be considered to define wedge shaped asperities (figure 5 -2 3).
Then if there is shear without rotation of j o int-blocks, the angles
io are defined fo r sli d ing in either direction along the mea n plane
and either equation (12) or equations (14) th ro ugh (17) can be
app l ied. Fi g ure 5- 23b sh ows a case in which over-sliding of asperi-
ties is pos sib le on ly in one d i recti on along the mean plane due to
the steep angle of imbrication; however, rotational sli din g can occur
in this case. Compoun d slide surfaces as in figure 5- 2 3 occur in
regu larly jointed rock masses sheared at intermediate orientations . .
However, whe r eas e q u a tion ( 12 ) and (14) require continuous wall rock,
a case wh ere one set of layers or bl ocks slides on another imp lies
a slidin g s y st em with additional degrees of freedom, t h e most impor-
tant of whic h relate to ind iv idual block rotations. The jamming of
a drawer when p ulled eccentrically illustrates the large modif ication
rotational tendencies can introduce. Nascimento and Texeira (1971)
d is c ussed rota tion a l f ri ction where shear oc curs on two parallel
cross - j oint e d sl id ing surfaces as in figure 5 - 2 4. In this figure
a continuous row of b lo cks, three of wh ich are s hown, is subjected
to left l at e ra l she a r by a force T applied at the upper platen.
The overtu r nin g moment of Tl on a block initiates rotation about the
left bottom corner (0) with the upper right corner (P) tending to
move alo n g a circular arc P pI, with dilatancy v. A small rotation
y may cause changes in N because of t he dilatancy and shear forces
develop a lon g the sides of the blocks, ult i mately inclining the side
f orces F2 a n d F3 at an angle ¢ to the normal with a back turn i n g
moment abo u t O. The displace ment a nd the rotation necessary to
neut ra lize t he mome n ts depend upon the magn itude of T a nd the stiff-
ness of the normal l o ad mechanism. Thus corresponding to every T,
up to a limit, there is an eq uilib ri um y . The limit is reached
when overtoppling occurs at y = 0 where 0 = tan-
l (a/b) ; instability

wi l l b e preceded by a fla tt en ing of the r at e of d ilatan c y , whi c h


becomes stati on a ry a t the pe ak lo ad , figure 5 -24b .
If the bl ocks b ecome separated so that side fo rc es are nil, t he
--1
v T I
-, ...J

I
I
I (a)
I
I b
I
I
I
I
0 a .1
!:::.v(t) dilatancy

o !:::.U [t)

(b) t

cr constant, low value

Figure 5-24. (a) Rotational friction; after Nascimento and Teixeira (1971). (b) Rotational shearing behavior
of bidilatant joints at low normal pressure; compare with figure 5-19.
Mechanical Properties of Discontinuities 195

line of action of Fl and its reacti on at 0 must l i e along t he dia-


gonal OP; in this cas e, the in c lination of Fl with the norm al l oad
is always ¢R = a-yo In the presenc e of side for c es , t h e rot at i onal
friction angle ¢R wil l be gre ater t han a - y. If Y i n c r eas es
uniformly with increasin g T, t h is i mpl i es that the ro t a ti ona l p eak
load variati on with 0 is convex downwa rd r esemb li ng e q u ation 14.
Accordingly it is n ot un re asonable to in clude the effe c ts of bl ock
rotation in t ha t e quation. Ladanyi an d Archambau lt (1972) did t hi s
through an adjus t ment of exponents Kl and K2 in e qu a tions (1 6a) a nd
(16b) as follows. Let n
be the number of r ows of bl o cks req uired
r
to defi n e the shearin g zone; f o r e xample, in figure 5- 24 , nr = 1 .
Results of shear tests wit h n in t h e rang e 2 to 5 could b e s imulated
r
by (14) th rough ( 16) by putting i n :

3
2
) t an i (1 9)
n o
r
and

(2 < nr < 5)

Reducing Kl makes the a rea of cont ac t, a sl smaller at an y normal


pressure, O. The in c r eased va lue of K2 i mp lies a f as te r los s of

Figure 5-25. Rupture modes observed in b iaxi al model tests with imbricated joint blocks; from Ladanvi and
Archambaul t (1972) . (a) format ion of a shear plane; () = 0; 0 3 = 28 Kg/ cm 2 (b) formation of a shear zone;
() = 30° ; 0 3 = 7 Kg/cm2 (c) formation of a kink ban d; () = 60°; 03 = 35 Kg/ cm 2
: 196 Mechanical Properties of Discontinuities

dilatancy with 0, associated with rock crushing along the edge to


face contacts. Reasonable agreement was found using this theory to
calculate results of biaxial model tests on imbricated block systems.
The blocks were of square cross-section 1.27 cm x 1.27 cm and 6.3 cm
long; 1800 were used per model. Three modes of failure were observed
depending on the orientation of the joint system:
(1) Shear along a well defined plane in general oblique to
both joint sets as in figure 5-25a; K2 = 4, Kl = 1.5;
(2) Formation of a narrow zone of rotated blocks with nr 2,
as in figure 5-25b; K2 5, Kl = tan io; and
(3) Failure along a wide band of rotated and separated columns
of blocks with nr = 3, as in figure 5-25c; K2 = 5, Kl = (2/3)3 tan io
The authors used the term "kink band" to describe this mode of
failure, which resembles kinking in metamorphic rocks as described
by Ramsay (1967), and Paterson and Weiss (1966). Kink bands occur
when sliding takes place on the discontinuous joints; movement of
any block will then exert an eccentric load on its neighbor which
overturns. This mechanism was called "block jacking" by Goodman
(1972). The failure process is similar to toppling of slopes de-
scribed by Hoek (1973), Goodman (1972), Cundall (1971), Barton (1971),
Ashby (1971) and others.
As an illustration of calculation by Ladanyi's theory, consider
the block model result of figure 5-26. The continuous joint set
0
was inclined 15 with respect to the direction of 01' Failure
occurred by formation of a kink band, two to four blocks wide, at
a mean orientation of ~ = 35 with the direction of 01' as shown
0

in figure 5-26. Instability occurred when the diagonal of the block


l
columns rotated into the direction of ° (i.e. y = 0 = tan- (1/3)
2
18 0 ). The confining pressure 03 = 13.5 kg/cm was applied all
2
around and then 01 was increased. At failure, 01 was 115 kg/cm
The normal stress, o, and shear stress T on the failure zone at
failure were, from (1)

2
T 47.7 kg/cm
P

2
(J 46.9 kg/cm
Mechanical Properties of Discontinuities 197

(a) (b)

(c) (d)

Figure 5-26. Development of a k ink band; 03 = 13.5 k g/cm 2 . Fro m Ladanyi and A rchambau lt ( 1972).

Corresp onding to a =
46.9, t he peak shear st r ess wi l l be c al c u -
2
l ate d b y equa ti on s (1 4 ) to (19 ) a n d c omp ared with 4 7 . 7 kg/cm . Figure
5-27a s h ows the r e lat i ve orientations of the two joint sets an d the
mean o rientation of th e k ink band . The orientation of the blocks
r e la tive to the d i re c ti o n of s h earin g is more readil y appreciated
fr om figur e 5- 27b in which the region of the eventual kink band is
s ho wn as if i n a d i r ect shea r test; the cont in uous j oints dip i n
t h e di rec t ion of th e s he a r l o a d or in t he negat i ve d irection a ccordin g
to t h e t e rm i n ol o g y of Ha yashi a nd Kiti h ara (1 970 ). It is easy to
app r eciate why block rot ations o ccur in such a tes t. Hayashi also
198 Mechanical Properties of Discontinuities

Figure 5-27. (a) Orientation of joints and


kink band (AA') in figure 5-26. (b) Orienta-
tion of blocks with respect to shear in direc- ~ 1
tion AA' of 5-27a.
A["""-__~_ _ _-r-_ _ _----'r---_-----.B

0; 0'3

-
---
---
mean
orientation
of kink band
(a)
OJ
0-

t l ~ l l
--. _1
----

(b)

found rotational modes in negative joint orientations giving reduced


shear strength and increased dilatancy, (see also figure 54 of Bernaix,
1974) .
Table 5-2 lists the parameters and results of calculation by
o
Ladanyi's theory. Using the actual value of 39 for ¢ of the model
II
material, the calculated shear strength T is too large. The actual
p 0
and calculated values of T agree for ¢ = 27. This low friction
p II
angle -- less than the residual friction angle -- is not unreasonable
Mechanical Properties of Discontinuities 199

TABLE 5-2
Calculation of Peak Strength
For Kink Band of Figure 5-27

Input

0 0 0
<P11 39 , 30 , 27
2
qu 254 Kg/cm
T 28.2 (n = 9 in equation 15)
0
0
i 40 (see figure 5-27)
0
n 3
r
K2 5
IT 46.9

Intermediate Results

a 0.057
s
v 0.261

Peak shear Strength compare with T p 47.7

71.
52.9
48.2

in view of Nascimento's discussion of rotational friction (figure


5-24) which showed a continuous decrease of <PR with shear strain.
The subject of rotational friction is very important in view of
the low shear strengths associ~ted with instability by buckling and
kinking. Unfortunately our appreciation and understanding of these
phenomena are just beginning.

The Role of Water

As previously discussed, water changes the chemistry of joint


surfaces and therefore modifies sliding friction values; the presence
of water also changes the consequences of wear during sliding. When
shear occurs quickly or drainage is slow, excess pore pressures will
develop in the wall rock by virtue of the stress changes during
200 • Mechanical Properties of Discontinuities

increasing shear. An additional increment of water pressure can


build up in the joint itself before peak stress if the volume of
the joint tends to contract. Rotational shear, in which large
dilatancy precedes peak load, should drain the excess water pressure
and raise the strength; but this has not yet been confirmed experi-
mentally. In non-rotational shear of smooth joints, Goodman and
Ohnishi (1973) observed continuous pore pressure rise until slip,
with a net loss in strength because of induced pore pressure; in
other words, dilatancy occurred too late to drain the pore pressure.
Handin and Stearns (1964) and others have demonstrated that the
effective stress principle is valid for rock joints in shear. Thus
we can write T
p
= f (0 - p) where p is the water pressure in the
joint; that is, the peak strength equations discussed earlier should
be considered to be effective stress criteria.

INFLUENCE OF JOINT ORIENTATION

A rock mass with parallel joints can support tensile stress in


some directions but not in others. Some principal stress orientations
are safe with respect to shear and others induce slip. Thus joints
introduce anisotropy in rock masses.
The joint opening and shear failure criteria previously con-
sidered refer to local coordinates normal and parallel to the plane
of the joint. Therefore they formulate global anisotropy implicitly,
since the peak shear strength can be expressed as a function of the
principal stresses and their direction through a transformation of
coordinates. It is instructive to consider such a transformation for
the very simple peak shear strength criterion:

(20)

Such an assumption simplifies the discussion but is not as


limiting as may at first appear since water pressure and cohesion
C can be introduced through the substitution
J

a (21)
Mechanical Properties of Discon t inu i t ies 201

where ° is the t ota l no r mal stre s s o n th e s h e ar p l a n e and p is t he


a
water pressure. Or , in te rms of princ ipal s t r e sses, substitute

CJ
°1 + p
°l,a tan <P
J
(22)
and
C
J
°3 ° 3,a + P
t an <P J

where ° 1 an d ° are the total p rincipal s t resses .


,a 3 ,a
Following Bray (196 7 ) , consi de r a d iscon tinu it y i ncli n ed ~

degrees wit h the d irection of ° , as in fi gure 5-3a. Then


1

° 1 + °3 °1 - °3
° c os 2 ljJ
2 2
and (2 3 )

°1 - °3
T s in 2 ljJ
2

Inserting equati o ns ( 23) i n ( 20), and int roduc in g Kf = ° 3,f I °1 ,f


where 03 f and ° f are the p ri ncip al st resses to p roduce s l ip:
, 1,

(1 + Kf ) t an ¢J - ( 1 - Kf ) cos 2 ~ tan <P


J
(24)

Consider ljJ p osi tive, so l v e for K a nd in troduce tri gon o met r ic i denti-
f
ties for sin (A + B ), a nd (s in A ± si n B ) to obtain t h e condition
for limiting equi li b rium:

t an ljJ
( 25)
(° 1 compressive )

This expre ssion is u nchan ged if i n p la c e of ~ we substitute


ljJl = 90° - (ljJ + ¢ J)' Fina l ly, if ljJ 2 = - ljJ i s i nt roduced, then i n
(24) only sin 2~ become s ne g ati ve and t he abso l ut e v alue brackets
restore the po s iti ve sign, s o that we a gain o btain t he resul t (2 5).
202 Mechanical Properties of Discontinuities

Thus we can conclude that four values of ~ (in each 180 0 range)
produce identical values of K , i.e.:
f

(26)

Equation (26) shows that the principal stress ratio to produce slip
on a joint with friction angle ~J varies markedly with the orientation
of the stresses with respect to the joint; the maximum value of Kf
(minimum strength) occurs when ~ = 45 - ~J/2.
0
This is shown in
Table 5-3, and in figures 5-28a and b, where K is plotted against
0
f
0
~ for ¢J = 20 and 40 respectively. Any set of values of 03' 01'
and ~ define a point by the polar coordinates (K = 03/01'~)'
Such a point is consistent with the criterion of slip only if it is
outside of the shaded region of the Bray diagram. During a joint
triaxial test, ~ is fixed and the test commences with K = 1. Then
as 01 - 03 is increased, K decreases and slip occurs when K = Kf .
Within the complete band - 90 < ~ < 90 0 , negative values of
K
f will be calculated from (21) when ~ > (90 - ~) or ~ < - (90 - ¢).
Since 01 is the algebraically greatest stress, with compression
positive, a negative value of K occurs when 03 is tensile. A certain
amount of tension is allowable as long as it is in a direction almost
parallel to the joint plane, i.e. 01 almost perpendicular to the
jOint plane. For a given orientation ~, slip will occur if
03 ~ 01 K f (see Table 5-3). A joint will open when cos 2~ =
(1 + K) / (1 - K). When both principal stresses are tensile, the
joint will open for all values of K.
The shear strength criterion for the rock can also be plotted
in the Bray diagram. For isotropic rock, Kf = f (03)' which can be
plotted as a series of concentric circles of radii f(03)' For example,
the Mohr-Coulomb criterion for the rock expressed in terms of princi-
pal stresses takes the form:

or

(27)
TA BLE 5-3
Li miting Principal Stress Ratio Kf = a 3 1a 1
For Slip on Jo ints in Different Orientations ( 1/J ) *

1JJ ¢: 20 ° 25 30 35 40 45 50 55 60

0 0 0 0 0 0 0 0 0 0

5 .1876 .1515 . 12 4 9 . 1043 .0875 . 0734 .06 1 3 . 0505 .0408


10 .30 5 4 . 2 5 18 . 2 10 1 .1 76 3 . 1480 . 1235 . 1 0 18 . 0822 . 06 42
15 . 3827 .3 19 3 .2679 . 22 4 8 . 1 876 . 1 547 . 12 4 9 .0 975 .0718
17 . 5 .0994
20 . 433 8 . 36 40 .30 54 .2 54 9 .2 10 1 . 1697 .1 32 5 .09 75 .0642
22 .5 . 1 7 16
25 . 4663 . 3913 .326 5 . 2692 .21 74 . 16 9 7 . 12 4 9 .0822 . 0 408
27. 5 .2 71 0
30 .4 845 . 4042 .33 3 3 .269 2 . 210 1 .1 54 7 .1018 .0505 0
32 . 5 . 4059
35 . 49 03 . 404 2 . 3265 . 254 9 . 1876 . 1235 .06 1 3 0 - .0613
40 . 4845 . 39 1 3 .3054 .22 4 8 . 1 4 80 .0 7 34 0 -.0734 -. 1 480
45 . 4663 . 364 0 .2679 . 1763 .0875 0 -. 0 875 -. 1763 -. 2679
50 . 4 3 38 .3193 . 2 101 . 104 3 0 -.1 0 4 3 -. 210 1 - . 319 3 -. 4 338
55 .3827 . 2 518 . 1249 0 -.1249 -. 25 1 8 -.3827 -.5198 -. 6660
60 . 3054 . 1 51 5 0 -.15 1 5 -.30 5 4 -. 464 1 -.6 304 -. 8077 - 1 .0000
65 . 1876 0 -.1 875 - .378 1 - .5746 - .7805 -1.0000 -1.238 1 -1 .50 16
70 0 -.2404 -.4845 -.7362 -1. 0000 -2.7475 -1.5863 -1.9238 -2.305 4
75 -.32 65 - . 65 81 - 1.0000 - 1.3584 -1.743 -2 . 15 4 7 -2.6132 -3.1315 -3.752 1
80 -1. 0 000 -1 .5 196 -2.0642 -2.6 446 -3.2743 -3.971 1 -4.7588 -5 . 6713 -6.7588
85 -3 .7321 -4 .1 6 0 2 -5.3299 -6. 5991 - 8.003 4 - 9 . 59 10 -1 1.430 1 -1 3.6218 - 16.3238
90 00 00 00 00 00 00 00 00

* 4J is t he angle between t h e plane of t he jOin t a nd t he di r ect ion of


°1
(s e e figur e 5-3a)
(a)
45-4/2)
~CTI

~~ ~\'#
1/1

\
\ CT
3
= KCT1

Mean plane
of joint

Locus
_ ton ~
Kf -tan(1/I+q,J)
Region of· ~
joint slip " '- __ 0 0 .1 0.2 0.3 0.4
I I I I I
K pas., 1/1 neg .
Scale for K = CT3 / OJ

-90< 1/1< 90°


Q)<K~I

(b)
As K is reduced at constant 1Jt:
failure IS on the joint at low 0"3 i
through the rock at high 0"3

--""~~::....:;::-...:>...::::"':::"'~~~"::::::""~~~~B-H------------...L--- Mean plane


of joint

o 0.1 0.2 0 .3 0.4


I I I I I

0
-90 <1Jt<90°
ll'<K~1

Figure 5-28. (a) Bray's diagram for the Coulomb criterion for joint slip for
<l>J = 20°; 0, compressive. (b) Bray's diagram for a joint with fPJ = 40° in rock with <I> = 50°.
Mechanical Properties of Discontinui ties 205

Thus as IT3 in c r ease s , Kf f o r the r o ck grows large r, as shown in Table


5-4 and in figure 5-28b where a cir c l e for ¢ = 50° has been supe r-
i mp os e d on t h e Bray dia g ram for ¢J = 40°. We can see from this figure
wh y , in c ert a i n ori ent a tio n s , the j o in t ceases to be weaker than the
r ock whe n 0 3 is a bo ve a g i ven v alue. F ig ure 5- 2 9 is an alternate
mode o f p re s e nt at ion in cartesi a n c o or d i nates x = ~, y = (01 - 0 3 )f
f or vary in g v a l ues of 0 3' Donath (19 64) , Jaeger ( 196 0 ), Chenevert
and Gatl in ( 1965) a n d others have plott ed anis ot r opy this way, while
Mu ller ( 19 6 6) and John ( 19 70 ) use a somewhat d iffere n t form with
po l a r c o o rd i n at e s ( (01 - 0 3 )f' ~).
Bray (196 7 ) used his graphi c a l presen t a t ion to discuss super-
position of slip crite r i a for multi - j oi n t ed rocks. He ingeniously
t r ace d the chang ing orientation and s i ze of t h e "petals" through
de velopment of imbr ication by slip on crossing joints. Of course

TABLE 5-4
Kf : Radius of Rock Failure in Bray D iagram

0 3 / Qu
Kf
2
1 + (0 3/ Qu) tan ( 4 5 + ¢/2)

40 0
0
0 3 /Q u ¢ ¢ 5 0° ¢ 60

0 0 0 0

.05 .0406 . 0 363 .0295

. 10 .0685 . 0 570 .0418

.2 0 .1042 .0 797 .0528


. 30 .1261 .09 19 .0579

. 40 . 1409 .099 5 .06 0 9

.50 .1 5 15 .1047 .06 28

.60 .15 9 6 . 1085 .0641

.70 .1660 . 1114 .0651

. 80 .171 0 .11 3 7 .0659


206 Mechanical Properties of Discontinuities

5.0

4.0

3.0
OJ -0"3
qu

2.0

0"3/Q u = 0.1
1.0

Figure 5-29. Anisotropy of a rock


with one joint: ¢J = 40°; ¢rock= 50°

superposition can afford only a first estimate for the shear strength
anisotropy of a rock mass cut by several sets of joints, since stress
redistributions during slip introduce local block breakage as well
as rotation and buckling. However, a closer approximation to the
real behavior of multi-jointed rock systems is possible with physical
model and finite element techniques, to be discussed later.
Although the discussion of strength anisotropy caused by a set
of joints has been two dimensional, it must be appreciated that a
peak stress criterion like (20) is truly three dimensional; for in
place of (23) we can introduce the three principal stresses in (20 )
through:

12 + 2 2
m + n
° °1 °2 °3
and (28)
2 m2 2 n 2 ) - 02
T j(012 12 + +
°
2 °3
Mechanical Properties of Discontinuities 207

N
peak strength
t
criterion for joi nt

o 03 c
A ,...
':~ ~ ~~
~ liJ:N i
50 L~ pn ~
r\. 1~'"
a rn
40 ~ K
I~
B
20

o 20 40
IV
50

Figure 5-30. Application of Mohr's circles to test for slip on joints of varying
orientation according to a given peak strength criterion.

where 1, m, and n are the cosines of the angles the normal to the
joint plane makes with the directions of 01' 02' and 03 respectively*
(figure 6-37b). A quick appreciation can be gained using the three
dimensional Mohr's circle diagram, figure 5-30. Given the directions
and magnitudes of 01' 02' and 03' a joint whose normal has direction
cosines 1, m, and n has normal and maximum shear stress defined by
the coordinates of point P determined as follows. From 02 lay
-1 1
off angles ~ = cos (n) and a = cos- (1) as shown in figure 5-30
defining points E and F. P is at the intersection of an arc through E
with center D and an arc through F with center C, where D =
1/2 (01 + 02) and C = 1/2 (02 + 03). If P defines a point in the

-1
In the two dimensional case considered, ~ corresponds to cos en).
*
208 Mechanical Properties of Discontinuities

ruled area, the stress and s.trength data are incompatible, i.e. slip
occurs. Using this construction, the locus AB of the peak strength
criterion across the Mohr's circle can be transformed to cartesian axes
(~, a) and the zone of safe joint orientations the Coulomb criterion,
were presented by Jaeger and Rosengren (1969). These show that the
relative magnitude of 02 exerts a significant influence on the relative
orientations of safe joints.
6
app Ica ions of
stereographic pro·ec ·0
in mechanics of
discon i 0 s rocks

INTRODUCT ION

In th e de si g n o f rock f oun d at ions and e x c avations, we become


con c ern ed wi th poten t i a l i nstabi li ty of rock blocks bounded by inter-
se cting planes of we akne s s . Th e st abil i t y analysis o f such blocks,
whe ther they are sub jected ma inl y to self we i ght , or to additional
loads such as static or dynam i c water forces, will usually begin with
a qualit a t ive inve sti ga tion to defi ne the mo r phology of the body and
t h e const ra in t s at th e limit of equil ibr i um, together with descript i on
of the mode of fail u r e. Afte r t hi s "kin e mati c ll
study , e quations fol'
stability a re p os tulat e d and the degree o f sa fe ty is calcu l ated. For
example, to de termine the possible l oads on the l i n ing of a large
tunnel it will be ne cessary fi rs t to de termine possible shapes o f wal l
or roof overbreak; the n, knowin g th e f r i ctiona l p roperties o f the
various r ock surf a c e s an d the un i t weigh t of the rock , i t will be
possible to specify a linin g to prov1de the p assive support necessary
to retain the r o ck in p lace. As an o ther il lustration , consider the
stability anal y si s f o r the ro ck a butme nts of an arch dam. The t hrust
of the dam may ten d to d isplace we dg e s a n d blocks, but countless
deformational mo des wi ll e xis t, each with di f f e rent possible e ffects
on t he dam . On e needs a k in ematical s tudy o f the syst e m o f discon-
t i n uities in re lati on to the dam in o rder t o kn ow whe re and how t o
draw IIfree bo dy di ag r a ms!'.
210 Applications of Stereographic Projection

This chapter introduces stereographic projection procedures for


both kinematical and equilibrium calculations in discontinuous rock
masses. The method of stereographic projection is selected for
simplicity, since the problems posed by any real case are invariably
three dimensional.
As discussed in chapter 3, stereographic projection relates to
operations on a unit reference sphere. In the applications to be
discussed, the reference sphere will be considered to be the envelope
of unit vectors radiating from a point. Operations with orientations
of vectors can be done directly on the stereonet, but most operations
relevant to statics demand an accounting of not only directions, but
magnitudes and positions of vectors as well. However, by combining
the stereographic projection with graphical or trigonometric opera-
tions, the necessary calculations can be completed. Thus the stereo-
graphic projection is useful not only in discussions of the shapes
of rock blocks formed by intersecting weakness surfaces, but in
analysis of stability of such blocks. Most of the examples discussed
in this section relate to the stability of blocks in excavations at
the ground surface. However, the problem of limit equilibrium of
one or more rock blocks is a basic stability problem for jointed rock
masses both at the surface and underground.

KINEMATICAL CONSIDERATIONS

"Kinematics" deals with the motion of bodies without reference


to the forces which move them. In the context of rock masses, this
implies the movement and modes of failure of the system of rock
blocks and joint surfaces. Jointing in hard rocks produces blocks
of varied shapes, some of which are illustrated in figure 6-1.
Regular patterns of equally spaced joints will produce cubes or
regular prisms. When the jointing directions are not perfectly
specified but rather are dispersed about preferred orientations,
tapered prisms arise. Slivers occur close to excavations where dis-
continuities lie at some small distance behind the free surface.
They will also be found in dispersed joint systems. In sedimentary
rocks, individual beds form flat slabs. The granitic rocks possess-
ing exfoliation jointing (sheeting), present families of slightly
Applications of Stereographic Projection 211

____CJ
L
cube regutar prism tapered prism sliver

L 7
flClt slab curved plat e tetrahedron

[] ~
~
hexagonal column square column incomplete tetrahedron incomplete pris m

Figure 6- 1. Some common block shapes.

curved sheets. Intersections of three discontin ui ty sets will pro duce


tetrahedra; four sets will produce truncated tetrahedra incomplete
tetrahedra or incomplete prisms. Finally, volcanic flow rocks
frequently have columnar structure formed by joints parallel to a
given axis. The u ndisturbed rock mass, as noted in Chapter 2, con-
sists of a tight packing of such blocks essentially without void
space (figure 6-2a). An excavation considerably larger than an
individual block liberates a large number of differently shaped
blocks which could potentially move into the excavation. It is
impractical to analyze each individual block independently, but
extreme combinations of members of each joint set can be considered.
Here, discussions of kinematics are particularly relevant. On the
other hand, one block alone is sometimes sufficiently large that it
merits individua l study as a single f eature. In such a case, t he
stereographic projection permits a rigorous discussion of force and
moment equilibrium.
Goodman- Geo l o9 'cal Eng -15
212 Applications of Stereographic Projection

contact

( a)

joints
;
.I .I
shear ./
"'"
,; v

I
I
- r \

"
contact ' "\.

, -" "
,
/ - I ...... ,
.t
'" , - .; '- I
\.
t \
\
, I J
"..... - \

(c) (d)

Figure 6-2. Examples of multiple blocks (a and c) and single blocks (b and d).

Lines of intersections between discontinuities give birth to a


prevalent failure mode of translation of rock blocks -- two planes
remain simultaneously in contact as translation occurs parallel to
the line of intersection, as shown in figure 6-3a. This is only
possible if the intersection "daylights" into a cutting, that is to
say, if it points out of the block into the free space (figure 6-3a).
A simple test of this requirement is possible on the stereographic
projection. First plot the orientation of the cutting S (figure 6-3b);
identify the daylighting portion of the lower hemisphere. Under
gravity alone, the slide is possible only if the intersection is
pointed downward at such an angle as to overcome friction on the two
o
planes. If the angle between the planes is close to 180 , a slide
will ensue whenever the line of intersection plunges more steeply
Applications of Stereographic Projection 213

( 0)
on intersection slide in 0 ( b)
rock cut region of dayl ight in g intersect ions

Figure 6-3. Sl iding along the intersectio n of d iscont inuit ies.

than the friction angle ¢ (assuming no cohesi o n ). In more a c ute


wedges, the li n e of intersection must pl un ge c onside r ably more steeply
than ¢ for sliding to be possible (Wittke, 1965 ) . As a simple
kinematic test, c omb i n e the two requirements, ( 1) for d ayli gh ting,
and (2) an intersection plunging steeper th a n ¢. A smal l c i r c le
90 - ¢ degrees about the vertical conta ins all lines plunging steepe r
than ¢. A daylighti n g intersection is in the area o utside the great
circle S. Superposition of these two kinemat ic tests yields the
ruled area shown i n figure 6-4a. Figu r e 6 - 4b offers an example.
Contours of orien t a tions of normals to joints a re s h own with the
preferred orientat i ons, i.e. the c en tral tendencies of t hree joints
as indicated by the symb o ls t;l' t; 2 and U ( f i g ure 6-4b ) . *
3
The line of interse c tion between t wo planes wi t h p oles 01 and u2 can
be found by const r ucting the n ormal to the gre a t ci rc le con t a ining
nl and n
2
(1 12 in figure 6 -4b). Of the three i nt erse ctions formed by
nl' n2 3
, on l y n and
and u 2 3
u
define an intersection in the p ot en-
tially unsafe region (1 in figure 6-4b ) . This con s t ruc tio n was
23
suggested by Markland (1972) and pub l ished by Hoek and Bray (1 9 74).
It is very us e ful fo r a qu i ck analy sis in c a s e s whe r e a g re a t many

* The symbol A over a letter s pec ifies a unit ve cto r.


214 Applications of Stereographic Projection

cut slope 5

(a) ( b)
LH LH

Figure 6-4. Simple tests for sliding along an intersection of discontinuities under
gravity loading only; after Hoek and Bray (1974).
(a) superposition of friction and daylighting tests;
(b) application of kinematic tests to a cut (S) in a rock mass.

possibilities exist.
Another construction can test the case of sliding on a single
plane, under gravity alone, because in this case the block must
translate down the dip of the plane (figur.e 6-5a). As shown in
figure 6-5b, for this mode to exist, the true dip must be contained

(a) [b) II"


Figure 6-5. Kinematics of a wedge sliding on a single plane under gravity loading
only; (a) gravity slide on one plal'!e; (b) kinematic requirement.
Applications of Stereographic Projection 2 15

5 ign conventIOn
i '2 d efined as lower hemisphere Intersection of planes 1 and 2
j32 defined as the Intersection of planes 3 and 2 havln9 a bearing
___---,r--w-it_hin the half circle of the bearings of -IJ5,Ds,Ic...
3s_ . - - -_ _

Kinematics
test 1 O2daylights and
steeper than 'f'
IS

test 2 O2 IS in smaller angle


betw een 11'2 a nd 137
101 Example dip vector ~ passes test 2 (bl Example DIp ve clor 02. passes both
but falls test 1; can not slide . tests ; can slide. D, falls both tests.
0, falls both tests.

Figure 6-6. Kinematic tests for wedge sliding on a single plane under gravity loadi ng only (see f igu re 6-5) .

between the two intersection lines limiting the e xtent of the pl ane
of sliding. A line of intersection will be considered p o siti ve if
it daylights into an excavation; figure 6-6 shows how to identify
the positive direction of an intersection for the example of '"132 i n
figure 6-5a. Form the intersection of plane 3 and the cut ting S;
the dip of the cutting is denoted DS' 132 is defined t h e n as the
intersection of planes 3 and 2 directed to have a be ar i ng within the
half circle formed by the lines 13s , US' -1 3s . The k i n e ma tic require-
ments for sliding down the dip of plane 2, for example, may then b e
stated in terms of the following two tests, both of wh i ch must be
passed in order that a slide occur. (1) D2 dayl ights and i s steeper
than the friction angle ~ and (2) D2 is in the small er angle between
112 and 132 , Figure 6-6a shows an example in which the di p vector
of plane 2 passes the second test but fails the first te st and
therefore cannot slide. In the same example the d i p vector o n plane
I fails both tests and therefore cannot slide. I n fig ure 6-6b, a n
example is given in which the dip vector on plane 2 passes both te st s
216 Applications of Stereographic Projection

( a) A blocked intersecti on

, --
~
"-4---
~
(2,
Tapered with respect to faces 1,2,3 Tapered with respect to faces 3,4

Figure 6-7. Tapering and blocking


Tapered with respect to faces 1,382,4 Tapered with respect to faces 1,3
of rock wedges; after Heuze and
Goodman (1972). (b) Tapered wedges

and therefore can slide. In this same example, fi l , the dip of plane
1, again fails both tests.
The above kinematic tests are useful for examination of large
numbers of blocks. When a single block is under study, further tests
can be introduced by tracing the relative locations of the normals
to the supporting planes. A block will be called tapered if it grows
wider as one traces its extent behind the excavation. Such a block
cannot by itself slide forward into an excavation. Figure 6-7b gives
four examples of tapered blocks. Figure Sa shows a five-sided block
with one free surface denoted S, which is tapered with respect to
faces 3 and 4. Such a block is tapered if the normals to opposing
faces, in this case 3 and 4, make an angle of less than lSOo, as
measured in the half space outside of the rock mass. Figure 6-Sb
shows how this test can be performed on the stereonet. Since n4 is
Applications of Stereographic Projection 217

Figure 6-8. A tapered block- kinemat ic test.

in t h e upper hemi sphere and ~ 3 is i n the lower hemisphere, the


op posite to ~4 has been sh own; therefore the required angle is the
suppl eme n t of that measu re d wi t h i n t h e l ower hemisp he re. A further
di scus sion of tape rin g can be fo und in He uz e and Goodman (1972 ).

OPERATION S WITH VECTORS ON THE STE REONET

The principal contribution of ste reographic projection in statics


lies in its simplifi c ation of ope r ations with vectors. Vector
operations -- the cross product, the dot product, addition a nd sub-
trac t ion, the triple scalar product and o the rs -- involve an apprecia-
tio n of the relati ve orient ations of the vectors involved. The
usual mathematic al method involves decomp os in g e ach vector into
orthogon a l comp onents in th ree basis directio ns chosen at will. But
t he st ereographic proje c tion enables the angles between the vectors
to be re ad d irectly a n d the refore an artifical decomposition onto an
ort hogo nal basis is n o l onger necessary.

The Cross Product of Two Vectors

Fi gu re 6-9 shows how the c ro ss pro du ct operat ion is pe rf o rmed


on the stereon et. Give n ve c tors A and B in di recti ons a a n d b. The
dir ecti ons of uni t vectors i an d b are shown on the stereographic
pro j e ction . Since t he di rec tio n of t he cross pro du ct of A and B is
p erpendicular to both, t he directi on of A x B lies in the direction
of the n ormal n, t o the pl an e common to a a nd b as shown in figure
218 Applications of Stereographic Projection

Figure 6-9. The Cross product of two vectors; sign convention-lower hemisphere.

6-9. The sign convention according to the righthand rule is that


C, the cross product of A and B is in the direction of advance of
a right hand screw when a is turned into b; the way this appears in
the stereographic projection depends upon the hemisphere convention
(figure 6-9).
The magnitude of A x B is equal to the magnitude of Atimes the
magnitude of B times the sine of the angle between a and b; this
angle can be read from a stereographic projection along the great
circle common to unit vectors a and h. If the vectors involved are
in different hemispheres, the operation can be performed with the
opposite to either ~ or b, reversing the order of the cross product,
since Ax B = B x (-A).
The Dot Product of Two Vectors

The dot product of two vectors A and B is given by the magnitude


of A times the magnitude of B times the cosine of the angle between
them. This angle again can be read in the great circle containing
the unit vectors of '"a and b.
'" Figure 6-10 gives an example.

Addition and Subtraction of Vectors

The addition of two vectors Aand B is formed by the parallelo-


Applications of Stereographic Projection 219

Figure 6-10. Multiplication of vectors-an example. N

,..
b

--+ --+ --+--+


FIND A X B and A· B
Given:
Magnitude IA I = 10 IBI = 25
Direction ais 30° down to N28°W
bis 21 ° down to N26°E
Solution to Cross Product:
--+ --+
Magnitude of IA X B 1= IAIIBI sin (J = (250) (0.765) = 192
--+ --+
Direction of AX B = 60° down to S200E
Solution to Dot Product
--+ --+ --+ --+ L. H.
A· B = IAIIBI cos (J = (250) (0.642) = 160

gram rule in the plane common to both vectors. To perform this


operation using the stereographic projection, plot the direction of
each unit vector and measure the angle between them; then on a
supplementary construction, as illustrated in figure 6-11, find the
resultant of the two, attaching the known magnitudes to the directions
derived from the stereographic projection. It is not necessary to
orient the vectors absolutely in the plane of the drawing, but only
to measure the angle between the unit vectors ~ and h. The resultant,
~ -r -+ A A

K, of A and B will lie along the great circle common to a and b, in


the smaller of the two angles between ~ an d b. In the case of vector
-+ -+
subtraction, the resultant A minus B will lie in the larger angle
between i and b along the great circle common to ~ and b (figure 6-11).
220 Applications of Stereographic Projection

Figure 6-". Addition and subtraction of vectors. North


~ ~ ~ ~

FIND A + B and A - B
Notation:
~ ~ /\ /\
A
~
= IAla = Aa
~ /\ /\

B = IBlb = Bb
~ ~ /\

A + B = Rl rl
~ ~ /\
A - B = R2 r2
Given: ~
A = 40 in direction
30° up to West
~

B = 20 in direction
70° up to South

- B B= 20

U.H.

Addition of Three Vectors by the Intersection of Great Circles

Londe, Vigier and Vormeringer (1970) showed a simple method of


summing three vectors on the stereographic projection (figure 6-12).
We seek the sum of vectors A, Band C. Since it is immaterial in
which order we add them, we arrive at the same answer if (1) we
first add 1 to B and then add C, or (2) we first add 1 to C and
subsequently add B. In the first case we will find the resultant Rab
in direction ~ b along the great circle between
~ a
i and b. When Ra b
is added to C, the resultant R b ' which we seek, will have direction
a c
~abc on the great circle between ~ab and ~, as shown in figure 6-12.
However, we can also proceed as in the second case by adding 1 to
Applications of Stereographic Projection 221

Notation :
A = Ao etc
A +B = Rob rob

-
A+ C Rae fcac
R = Robe r obe
r--.
a b means plane co mmon to a and b

(p,) td p,) mea ns li ne of intersection of


p lanes PI and P'2
Cons truction f or rabc

Figure 6-12. A ddition of 3 vectors by the intersection of great circl es.

C with
~
the resultant in the direction r
~ac
. Adding Ra c to
~
B fixes
rabc along the great circle common to rac a nd B. Th us rabc must
lie in the intersection of the great circles common to r ab and c,
and r a n d b (figur e 6-12). A form u l a for this cons tr uc t io n can be
ac ~
written by introduc ing notation as follows: le t PQ mean the g reat
circle common to lines P and Q; and let nJ\m signify t he line of
intersection of planes nand m. Then the unit vect or r b de fi n ing
a c
the direction of A + E + C is
~/\~
r ab c r ac 0 (1)

The above construction is useful where t wo of the vec tors i n-


volved are not known absolutely and are to be left as v ariables i n
an analysis. In the example of figure 6-13, vec to r A is con s idered
known while vect o r s Band Car e left as free parameters. Si de con-
struction gives the angle of rotation from a co rre sp onding to addi ng
r-.
given amounts of B; - thus the great circle ar-....b is "calib rated" with
values of lEI (figure 6-13). Great circle a c is simi l a rly c al ibra-
ted. Once these great circles have been labelled, the direct ion of
the resultant of A plus any combination of Band C can b e dete rmined
222 Applications of Stereographic Projection

Figure 6-13. Parametric addition of 3 vectors.

Given:
-+
A = Aa
b,2
Find: "/
/
-+
75° 1.5 /
D = Aa" + kb b" + kc "C 1.25 /
for all values of kb and k c ' k IA
b .75
to /

Solution illustrated for: .5 ~/


1,\
/
-A + .75 A b + • 25 A C
"
kb = 0.75 A
kc = 0.25 A II \
/ \
/ \
\
\
\
\
\
\
\
\
A
\
\
\
\

by the intersection of great circles, according to the method of


figure 6-12, without any further side constructions. This approach
is useful in analyzing the stability of blocks of rock which are
loaded by water pressure forces on the faces, as for example in the
abutments of a dam, in which the magnitudes of water forces will not
be known very well during design and will change as the reservoir
changes elevation.

The Decomposition of a Vector into Components

Given three basis directions b1 , b2 and b3 , any vector R can be


Applications of Stereographic Projection 223

North Figure 6-14_ Decom positio n of a vector into


non-orthogonal com ponents.

Find:
--+
Components of R = Rf
in basis directions bi ' b2 , b3
Solution:
BI , B2 , B3

decomposed into components along them. It is not nece s sa ry f or the


basis to be orthogonal. The method is illustrated in figure 6 - 14,
where r is the direction of R. The angle 8
3
from ~ to b3 can be
measured in the great circle common to
A A
b3 and r. Also construct
A~~
the great circle common to b 1 and b 2 . The great circles b I D2
and b"
r intersect at r
" '"
; read angle 8
A
between r and r. Now, " "

3 12 12 12
knowing the angles 8 and 8 , construct a triangle of for ce s to
-+ -+
3 12 " -+
decompose R into B 3 , (the component in " the di~ect i o n b 3 ) a n d R 12 ,
(the component in the plane common to b and b ). Then reading the
/"\, /\. /"\ l 2 +
angles a 1 and a 2 from r to b and b r es p ect i ve ly, dec omp o se R12
-+
12 -+ 1 2 -+ -+
into the components Bl and B2 , as shown in figure 6 - 14. Bl , B2 and
-+
B3 are the required components.
224 Applications of Stereographlc Projection

Figure 6-15. The largest tetrahedral wedge North


defined by two sets of discontinuities.

(15 oj

c
r
/
/
----------:::;oo~ :--\---100m /
edge of 2 /1\,
/ /. . . 1,2
115 ........ 2LgO 11$ /
"'" 38°..... 30 m
B~~------------~r-----~p-90m // section q, q'2
_----=:~---_-___,__---7~_12.L 80m ~'f
--~~,-----------+-~""---- \ 70 m

------~~------ 60m
A
L.. q
2
IABI = 30/si n 29° = 61 · 8
IACI = 42/sin 40° = 65'4
o 20 40 IADI = 30/sln 38° = 48·8
(15 b)
Applications of Stereographic Projection 225

APP LICATION O F THE STEREOG RAPH IC PR OJ ECTION IN DEF IN ING


A WEDGE FO RMED BY INTERSECTIN G DISCONTINUITI ES

Figure 6-l5a presents a h ypothetical example of a rock mass


contain ing a system of jO in ts and be ds in va ri ous directions. A cut
to b e e xc avate d parallel to plane S wi ll produce dayli g hti ng inter-
se ctions of t h e j oints and beds as shown in figure 6-l5b. It can
be shown t h at in t h e case of a material wh ich has both frictional
and cohe s ive strengths, the most c r itical member of a set of similar
unblo cked wedges wi ll be the largest of the set. If the cutting has
a specif ic hei ght, t h is wedge is unique ly defined; thus even though
there are mu ltiple similar blo ck s formed by t h e in tersecti on s of
t h e joints and beds in this example, for the case of limiting
equilibrium only one needs to be analyzed.
In fi g ure 6-15b the surfac e of t h e c ut is S, and the original
slope, wh i ch strikes parallel to S , is ca lled plane 3. (The solution
would be the same if plane 3 were the underside of another block
Sitting on top of the wedge. ) From the stereographic projectio n,
determi ne the i nt e r sections of the j oi nts an d the beds (112 ) an d
t h e i nt ersections of ea ch of these discontinuities on the cutti ng
(l lS) and(I 2S )· A In the plan view (left ha lf of fi g ure 6 - 15b)
the be a rings of I and 1 can be laid off f rom a verte x A c hos e n
1S 2S
anywhere along t he b ase of the 30 meter high excavation. Also
con struct the b e ar in g of the line of intersection 112 ; then, from
t he stereographic proj e ct ion , determi n e t he i n tersections 113 and 123
of planes 1 an d 2 wi th p lane 3, the top of the slope surface; laying
off the bearings of 113 and 123 f rom points Band D completes the
p lan view of the we dge. As sh own in fi gure 6 - l 5b, the magnitudes
~ ~ .~

of the vectors AB, AC and AD can be computed f rom the scaled lengths
on the dra wings and from the kn own plunge a ngles listed in figure
6 - l5b. To check th e accura cy of t h e construction, o n e can also
determine po int C b y constructing a vertical section in plane q
perpendicu lar to the c u t as sh own in the right half of figure 6 - 15b.
In this secti on one does n ot see the true p lun ge of t he line of
A 1
inte r section 112 b ut rathe r it s ort ho g raph ic pro jection 112 in th e
pl an e of the secti o n (the orth og raphic proje c tion of a line in a
pl a ne was shown i n figu r e 3-1 2) .
226 Applications of Stereographic Projection

Figure 6-16. (a) Volume and area of faces North


for the wedge of Hgure 6-15. (b) The
centroid of the wedge of figure 6-15.

./"-...
1\ 1\
ICAxDA)
0-. ....
.... .... _Or. ,
-<.::. ......60 0

'-'-, 1'5
'f-> (B"A)
I
I
I
/
/
/
I
o /
4\\0-
/

12s //

-~_~~_2_4~~ __
(DA) 112

((A)

(a)

We will need to determine the volume of the wedge ABCD and the
area of its contacting faces ABC and ACD. The volume of a tetra-
hedron can be calculated with the scalar triple product of any three
~

edges radiating from a given corner. Since we know the vectors AB,
-----4- ~

AC and AD, we can therefore determine the volume:

1 ~ ----+ .~
v 6" AB AC x AD (2)

as shown in figure 6-l6a. The operation of the scalar triple product


given above is a combination of operations previously defined; it
Applications of Stereographic Projection 227

North

Centroid

AM = _1t. [ 8A~ CA ~ DA]


= -33 W
( IN as show n on
stereogr am )

j~s 1,'2
DA 1'\
CA

(b)
L.H.

will be necessary only to determine the angles in vol v ed for th e d o t


product and the cross product. In this example, the we dge is 6,700
cu. meters in size ( know ing the unit weight o f t h e r o ck, its wei gh t
is also determi ne d ). The areas of the f a ces ABC an d ADC a r e c a lc u-
lated by half the ma gn itude of the cross pro duct o f v e c t ors along
the edges of the two faces as g iven in figure 6-l6 a. Fo r e x a mp le,
~ ~

the area ABC is 1/ 2 I AB x AC I. If the cohesion along the f ace is


known, the maximum c oh e sive force on ea c h f a ce can b e de termi ne d from
the known a re as. The centroid of the tetrahedral we dge c an be
determined by a vector addition accordi ng to the f ormu l a
Goodman-Geo lO<)lcal Eng.-l b
228 Applications of Stereographic Projection

-+ 1
AM (3 )
4"

hedron.
-
where AM is the vector from corner A to the centroid of the tetra-
The sum of three vectors was discussed previously. In
figure 6-16b, the summation is done by a repeated operation of the
summation of two vectors; the great circle method discussed earlier
for summation of three vectors is not necessary here since all three
vectors are known absolutely.

Position Vectors to Fix the Line of Action of a Force

Until now we have considered only free vectors. To fix the


position of a vector, one can describe a family of "position vectors"
radiating from a known pOint, whose tips all lie along the desired
-r
line of action. In figure 6-17a: force F is known in magnitude and
direction; point 0 is a known point near the region of interest; and
A is a point known to be on the line of action of force F. Then the
vector oA will have its tip in the line of action of force F, as will
the vector sum 01 + KF where K is any positive or negative number.
---* -r
The family of vectors OA + KF can be considered the vector description
of the line of action of F.
The Vector Equation of a Plane

A family of position vectors can also establish the position of


aA plane (figure 6-17b). Consider plane p with
_
normal
A
in direction
np For any point C in plane P, the scalar OC n p is constant. This
constant is determined by any known point C , in the plane, giving
l
as the equation of the plane

-+
OC n n (4)
p p

The Piercing Point of a Force on a Plane

By equating the line of act Lon of a force and the equation of


a plane, one determines the particular position vector whose tip
lies at the piercing point of the force on the plane. Figure 6-18
presents an example and shows how the resulting vector equation can
Applications of Stereographic Projection 229

--+
Line of action of force F is defined by the locus o
--+ --+
of arrows of the family of vectors OA + KF
where 0 is a known referen ce point and A is any
--+
point on the line of action of force F , (for example
its point of application).

o
(K= 0)

( b)

Position of plane p is defined by the locus


-+
of arrows of the fam ily of vectors OC
--+ /\
defined by the equation OC . np = constant .
/ The value of the constant is obtained if any
/
(01
/ point in the plane (C 1 ) is known.
I

Figure 6-17. Position vectors defining the line of act ion of a force and the equati on of a plane.

be solved iteratively on the stereonet. In this examp le we wish


to find the point whe re the weig h t of the wedge of fi g u re 6- 15
pierces plane 1. The line of action of the weight force p a sses
thro ugh the centroi d M whose position is known with respe c t to poi nt
A of the wedge, fi g u r e 6 -l6b. Let point D of figure 6-1 5b se r v e a s
-----+
reference. The ve ct or DM can be found by vector subtraction, as
--+ -;:::--t -----+
shown in figure 6 - 1 Sb , DM = DA - MA. We denote the ve rt ical ly
A

downwa r d direction as Z; then the equation of the li ne of action of


-----+ A

the weight force is DM + KZ. We k n ow th e location of p o int A in


the plane so the constant of the equation of plane 1 is d ete rm i n e d
(constan t = DA . ~l)'
-----+
The piercing point T on the p l an e is g i v e n
b y a position vector DT found by the value of K g iving a so lut ion to

-----+ '" -----+


(DM + KZ) . n DA (5)
l
230 Applications of Stereographic Projection

Figure 6-18. (a) Stereographic projection North


for iterative solution of a vector equation.
(b) Construction for the position vector to
the piercing point of a force; compare with
figure 6-15 and 6-16.

Figure 180

This equation can be solved iteratively on the stereographic pro-


jection together with a side construction. For every value K, the
-----jo- /'\

resultant DM + KZ will lie somewhere along the great circle between


~ /'\

the directions of DM and Z, as shown in figure 6-18a. At each of


~
a number of points along this great circle, the angles to DM and to
/'\

n 1 are read from the stereonet and plotted as shown in the upper
right of figure 6-18b. The dot product of two vectors is a scalar
quantity; it is shown by a circle of appropriate radius about the
point D in the lower left of figure 6-1Sb. In this example, the
value DA . ~ is equal to 10.2. For a value of K = 0 we can draw
the vector DM as shown in the lower left of figure 6-18b and form its
dot product so as to produce a value of 10.2 by drawing a tangent to
Applications of Stereographic Projection 23 1

I
ygon
"- ~ Fore e pel
~

""
I

r--!-
I ..........
I
I
10(
I'--- from st ereonet
I ' 5 (fig 180)
DA = 1.9 DA 1 I
o 10 20 30 40 50
0< ,- Angle along D~M-Z

K=30 Iterative Solu tion


Piercing point (T) of weigh t force on plane
-+ J\. -+ /\ /\ -+ /\
DT • n 1 = (DM + KZ) . n 1 = DA . n 1
-+
IDAI = 49 ; < DA to 111 = 78 °
-+ /\
DA . n 1 = 49 cos 78 = 10.2

From force polygon and stereonet,

0'1 = 5° giving k = 3.0.


-+ /\
Figure 18b IDT I = 33; DT shown on stereogram.

the scalar circle from point M, meeting the circle at p o int 1.


Corresponding to the choice a l = °
that we have made in taking the
value of K = 0, we have found the angle MD I (angle a 2 ') eq u al to
72 0 . Corresponding to each assumption of aI' there will be another
value of a ' and these can be plotted as in the upper right of
2
figure 6-18b. The intersection of the two curves in the upper rig h t
figure of 6-18b defines the value of a and therefore the value of Kj
l
K is equal to 3.0 and a equals 50. The vector DT can now be plot t ed
l
on the stereographic projection.

ANALYSIS OF ROlATI ON

The previous operations with vectors permit analysis of rota-


tional tendencies in rock wedges. A wedge may be a b le to topp le
about its edges or rotate while sliding in a given face , as discussed
by Wittke (1965), Goodman and Taylor (1967) and Londe, Vi gi er a nd
Vormeringer (1969). In this section we will explore the operations
necessary to analyze the general problem of rotation about an axis.
232 Applications of Stereographic Projection

Figure 6-19. (a) Rotation in a plane. (b) Analysis of rotation in a plane (see figure 6-19a).
(c) Stereographic projection for analysis of rotation.

(a)

~,>- ":-free space ' -


(b)

----------05
//8 ,
plane 2

Kinematics of rotation about A


~I
I
Angle between edges of face p, measured in plane p, must
!l>90° be greater than 90° for rotation (See figure 19c).
~ rotation requires punching N > 90°
I~ - kinematically improbable
f case (1) "l 90°
Analysis of rotation about A
(viewed 1 p) Axis through A 1 to plane p is in direction ~p.
~

T is piercing point of resultant R on p. Its location is


~

given by vector AT.


~
Overturning moment MA caused by R, which acts at T is
~ ~1\
AT X R • np
Maximum resisting moment MR depends upon the stress
distribution. Assume resisting force To is concentrated
at T,
~

T 0rnax (R· ~p) tan ~F


~

MR = To • IATI
~ >90
pI Stability requires MR > MA
A pplications of Stereograph ic Projection 233

In the case o f a f r ee b lock , a t r an s l a ti o n a l move men t can be v i e we d


a s a rot a t i on a b o ut a n a x is dire cted pe r p e ndicu lar t o the plane of
sli din g a nd lo c at e d a t i n f i nit y ; t he r efo re, r o t a t ion about any f inite
lever a rm mus t b e l es s cr iti c a l t h a n t r a n s lat i on . Howev e r, i f the
wedge is not unif o r m i n its p r op ert i e s , f or e xample if it has
a nis ot r o pic she ar s t r en g t h pro p e r ti es o r i f i t i s r e s tricte d i n i ts
transla t i on i n p a r t i c u l a r d ire c t i on s , th en r o ta t iona l slidi n g may
become cr i ti c a l.
Con side r t he we dg e of f igure 6- 19 a wi t h r o tat i o nal t ende ncy
a b out the co rn e r A. Th e e qu a ti on fo r t h e mo me nt o f th e force R about
an a x is in d i r e c t ion dA th roug h p oi n t A is g i v e n b y MA = AT x R . dA
whe r e T is a po i n t o n t h e lin e of a c t ion of R. I f the momen t is
positive it me a n s t h at t h e se n s e of r otat i on i s c l ockwi se a s v'ewed
234 Applications of Stereographic Projection

from the tail of the vector dA . We will examine the moment for the
rotational sliding mode of figure 6-19a, that is rotation about an
axis (~p) perpendicular to plane p through point A. Force R pierces
plane p at point T. The moment about the axis shown, MA , is equal to
"AT x R . ~p' If np is pointed into the lower hemisphere, as in this
example, then a negative moment corresponds to kinematically possible
rotation. If we view the plane of rotational sliding as in figure
6-19b, it will be seen that rotation about A is possible only if the
angle n shown is greater than 90 0 ; n is the angle between the inter-
A A

sections I 1p and I 2p ' as shown in figure 6-19c. If n is less than


0
90 , rotation is possible about an instantaneous center located at
the intersection of perpendiculars to plane. 2 and plane I drawn
from the points of contact, as discussed by Wittke, and Londe et al;
this case will not be considered here. The driving moment about the
axis perpendicular to plane p through A can be calculated from the
formula presented above. The resisting moment which must be compared
to the overturning moment is more difficult to discuss, as it depends
upon the stress distribution in the plane. Dr. John Bray* showed
that the most critical condition can be found by assuming that the
entire resisting force in plane p is mobilized at the piercing point
(T) of the resultant force R on plane p. In this case, the resisting
moment MR o times distance IATI; T 0 is equal to R • p tan ¢p .
= T n
Stability require~ that MR be greater than MA and therefore the
frictional requirements for equilibrium for overturning can be
examined.
The method presented above only makes sense if the piercing
point of the force R intersects the plane of rotational sliding on
one of the faces of the wedge. It is convenient to examine this
requirement using the stereographic projection. Figures 6-19c and
20a present alternative methods of doing this. With respect to
figure 6-19b, the piercing point T will lie in plane p if
A A A
AT is
between I and -I and the vector TB lies between the intersec-
sp 2p~
tions I and 1 , TB can be found by vector subtraction:
-+ ~ .> ----+ 1 p
TB AB - AT. The method of figure 6-20a can also be used to

* Personal communication.
(a) North c
t

LH

(b)

Figure 6-20. (a) Example of stereograph ic projection to test if a f orce pierces a


plane within a wedge. (b) Vector subtraction for the stereographic projection of
figure 5-20a.
236 Applications of Stereographic Projection

satisfy that the piercing point of T on plane 1 is on the face of


the wedge. By vector subtraction, form the vectors DA, DB and DC
~
as well as the vector DT to the piercing point from D. The unit
----+ ~
vectors representing the directions of the edge vectors DA, DB and
~
DC form three points on the stereographic projection which, when
connected by great circles, define a spherical triangle. If the
~

direction of DT is inside this spherical triangle, the piercing point


is on the face of the wedge. In this figure, the piercing point of
the weight force has been determined as discussed in figure 6-15,
and is found to intersect plane 1 of the wedge.
If one were to add another force (to the problem of figure 6-15)
which did not happen to be coplanar with the weight force, it would
not be possible to form a single resultant to the applied forces on
the wedge. Instead, first determine the piercing point of each force
in turn (points P and Q in figure 6-21); then, by taking the dot
product of each force with the normal to the plane in question,
calculate the contribution of each force to the total normal force
on the plane. In this example, the force P1 pierces plane 1 at P
while force F2 pierces plane 1 at Q. The normal forces produced by
± + + +
Y and F2 are Nl and N2 respectively.
By taking moments about any
l
point in the plane, for example point A, we will find the distance

Rotation analysIs with 2 forces

c View .1. plane 1

Section Figure 6-21. Analysis of rotation with multiple forces.


Applications of Stereograph ic Projection 237

r from A a t which 1+Nl + +N2 I acts with the same moment as the sum of
+ +
Nl and N2 ·

r (6 )

where r and r de scribe the location of P and Q, fi gu re 6-2 1 . The


1 2
most critical stress distribution of the resisting for c e on plane 1
corresponds to the case where the entire frictional resis t a nce T is
o
concentrated a t a radial distance r from A. (It does n o t mat te r in
what directi on as lon g as it is in t h e face ACB) If MA is the com-
+ +
bined moment of Fl a n d F2 about A, determined by re p eated a p pli cati o n
of the moment e qua tio n , then the condition for safety is:

T ( 7)
o

where ¢ aval"I a bl e is the friction angle available on plane 1.

ANALYSIS OF SLI DING OF A BLOC K


ON A PLANE- THE FR ICT ION CIRCLE CONCEPT

Under pure frictio n, the limit of stati c equilibri um of a b lo ck


resting on a plane o c c u rs when the res ul tan t force of the b lock on
the plane becomes inclined at angle ¢ from the normal directed out
of the block into t he plane. Therefore a block free to move i n any
direction but which is actually at rest on a plane p, must have a
resultant force o r iented inside a circular cone of vertex ang le 2¢
centere d a bout the n ormal to p (figure 6 - 22) , Plott i ng th is "con e
of static fr i c ti o n " on the stereographic projection produces a c i r c l e
"-
about the n o rmal n p ' as shown in figure 6 -22b . The po r ti on of t he
wh ole sp here inside t he fri ct ion circle wi ll be called th e " s afe re-
g ion " . Outside of this circle it is not possible for a re sultan t to
find an equilibri um reaction (although it may be convenient to disc u ss
the orientat i ons of potential forces outsi d e of the frict i o n ci rc le ).
The condit i on of l imitin g equ i librium i s f o und wh e n the direc t i on
of the resul t a n t force of the hlock on its support lies exact l y on
238 Applications of Stereographic Projection

(a)

L.H.
The friction cone about the normal to Stereographic projection of friction
an inclined plane surface of sliding. cone producing a "friction circle"
If the block is at rest. the resultant
force between block and plane is
inside the cone tp is the "friction angle"

(b) np is the central tendency


among the scattered joint
atti tudes

safe zone.
anisotropic
friction
roughness
'zone'

LH LH

Figure 6-22. (a) The friction circle concept. (b) Generalized friction circle for rough planes.

the circumference of the friction circle. Though simple, this concept


is powerful and, as will be shown later, can be generalized for
blocks resting on several planes. The basic idea involved is that
we sub-divide the whole sphere into a safe region and an unsafe
Applications of Stereographic Projection 239

region. The safe region is circular for isotropic friction on a


plane. It will be non-circular if the structures in the plane develop
anisotropic frictional resistance. As discussed in chapter 5,
repeated measurements of the attitude of a rock plane made on a flat
plate of given diameter will develop a scatter diagram of poles t h a t
can be interpreted to give the roughness angle on the surface in
question. If the generally non-circular roughness figure is added
to the residual friction value ¢, it will produce a non-circular safe
zone as shown in figure 6-22b. In this illustration, it was necessary
to pick a central tendency from within the distribution of p oles,
presuming it to be a unimodal pole distribution (chapter 3 ).
While it is convenient to carry out the stability an a lysis for
sliding on a discontinuity having the orientation defined by a si ngle
normal (~ ), discontinuities are actually scattered about this
p
orientation, as discussed in chapter 3. However, it is possible to
account for uncertainty of orientation using a reduced friction circle.
Suppose that the distribution of normals is such that a probability
of occurrence equal to P determines an angle of uncert a inty ~(P) as
in figure 3-16. On figure 6-23, a small circle of radius ~ has been

.p (Pl

o~--------------------~
99%
.. p

Figure 6-23. The relationship between the friction angle to be used with the mean
joint orientation and the probability of occurrence P.
240 Applications of Stereographic Projection

drawn about np , which, for convenience, has been rotated to the


center of the lower hemisphere. This small circle, which shall be
called the "probability circle" contains (P) x (100%) of the normals
to discontinuities of the set being analyzed. We will shrink the
safe zone about n p to insure that it is contained in the friction
circle of every normal inside the probability circle. Construct a
family of friction circles about points on the circumference of the
probability circle; the area common to the entire family is a circle
of radius ¢ -~. Thus, if we refer the construction of friction
circles to the preferred orientation of the set of discontinuities,
we must reduce the friction angle to:

¢(P) ¢available - ~(P) (8)

This conclusion may seem to contradict the construction presented


in figure 6-22b, but in that case we were dealing with certain
departures from a mean planar orientation for a rigid contacting
plate; in figure 6-23, on the other hand, we are concerned with an
uncertainty in the actual orientation of the plane itself. It may
still be desired to introduce roughness into the probability circle,
but then the area common to all the friction circles one can draw
within the cone of uncertainty would be rather complex.
What is the relationship between probability of occurrence (P)
and the probability of safety (PS)? If the direction of the re-
sultant force on a potentially sliding block plots on the circum-
ference of a friction circle of radius ¢ aval'1 a bl e constructed about
~ , the probability of safety is one half because half the joints are
P A
flatter than the orientation of np . The probability of occurrence
(P) that defines such a friction circle about n p is zero. Therefore,
the probability of safety and the probability of occurrence are
related by P 2PS - 1.0, giving for PS > 0.50:

¢(PS) ~(2PS - 1.0) (8a)


¢available

Applications of the Friction Circle

To illustrate applications of the friction circle concept,


various stability calculations will be made with the example of figure
Applications of Stereographic Projection 241

6 - 24a in wh ich p l a ne p da y l ight s i nto an e xcava tion, with a p ote n-


tially sliding blo ck we i g h i n g 4,000 t o ns. Th e area of the face of
the block par alle l to p is 200 sq. met ers , and t he f riction angle
(¢availab le) is 300 . Uncerta int y i n t he att it ude o f plane p can b e
rep re s e n t e d by a no r ma l di s tribution a b out the att i t ude given with
K 12 0 ( s e e c hap t e r 3). We fi rst e xami n e t he condition of
st ab il ity f or t he b lock un de r its own we igh t (W) b y cons t ructin g a
ci r cle o f radius 300 about ~.
p
S i nce t he weigh t p lots i n the center
o f the lower hemi s p he r e, it lies ou tside of the friction c i rcle;
he nce t h e b lock is unstabl e if i ts wei ght is th e s al e contribution
to the re sul tant f orce bet wee n th e blo ck an d plane p . Rock b o lt s can
be anc ho re d be low t he s l i p sur f ace to ach i eve a desired factor of
s a fe ty ( FS ) . The f a c tor of sa fe ty wi ll b e d e fined as t he ratio of
th e avai lab le fri c t ion to th e f riction require d f or l imiting equ i l i b-
rium under the given fo rc e s:

tan <P available


FS (9)
tan <P require d

A sa f ety f acto r of 1 .0 c o rres ponds to the con d i t ion where the r es ul-
tant f o rce plots on t h e ci rc umfe ren ce o f t he fric ti on circle having
0
radius ¢ = 30 . To achi eve th is condi t ion, a f orce can be added to
Wto inc l i ne t h e r esu lt ant 20 0 f r om ve rtical. The magnitude of the
r ock bolt forc e re q uired to do t h i s depends upon the orientation of
the bo lt s. The r e quire d force is sm a ll e st if i t i s directed per-
pendic u la r ly to the re sultan t, t hat is, 20° abo ve the horizontal as
measure d a long t he d i r e cti on of t h e bearing of np (figure 6-24b).
Since t h is direction is in t h e upper h e mis phere, its opposite
( - B l . O) has b een s hown .
Say a fa cto r of sa fe ty of 1 .7 is des i r e d. This means that the
require d f r i c t i o n angl e (<P r eq ulr
. e d) is equal to arc tan
0
(tan <Pavai l ab l e / 1 .7) or 1 8 .5 • To achieve thi s th e rock bolt force
0
must inc line t he re su ltant 31.5 ab ove t he vert ica l alon g the
di rect ion of beari ng ~.
p
The p o lygon of f orces pres e nted in figure
6- 24b, l e ft, determines t he r e qu ired r ock bo lt forces as 1 ,400 tons
a n d 2, 100 t on s r e spe cti ve ly f o r f ac t ors of s a f e ty of 1 .0 and 1.7.
242 Applications of Stereographic Projection

(a) N

_ 0

..,.
W ~(J

00-899%
0~B1.7
-B
10
Opposites to bolting directions
. also opposite to worst acceleration
direction
Figure 6-24. (al Example
of stability analysis.
(b) Completion of stability
analysis.

Though this is the minimum force, and therefore minimizes the required
cross-sectional area of steel, it may not be the most economical
direction in which to emplace rock bolts in a particular problem;
the geometry of the block may dictate other directions for minimum
length of drilling necessary to seat the anchorage of the rock bolts
well into the support. Every specific case will require the determin-
ation of an orientation optimizing the economic factors related to
the area and expense of steel on the one hand, and the length and
cost of drilling on the other.
Now let us design the rock bolts for a given probability of
safety. To design for a probability of safety (PS) equal to 0.99,
for example, compute Wcorresponding to P = 0.98. With K = 120,
equation 2 in chapter 3 will give W(.98) equal to ISo; that is, 98%
Applications of Stereographic Projection 243

(b)

Vi

W = 4000 tons
A = 200 m2
I .. ,
100 0 tons

(1) and (2)


Minimum rock bolt force Required cohesion (CIA )
FS = 1. 0 1400 ton s 8·0 tons 1m2
FS 1· 7 210 0 11 ·1
prob. safety 99 % 2 300 12 · 0

(I.) t he water pressure to initiate slip from an init i a l factor of safety


of 1· 7 = UIA = 5 ton s /m 2
(5 ) the smallest acceler at ion Kyg to initiate slip from an initial tactor
of safety of 1·7 is directed as shown by the vector KyW ; Ky=O.ll

o f the poles wi ll b e oriented wi thin 15 0 of the direction n.


p
There-
fore, to have 99% probability of safety, con s tr uct the f ri c tion
circle with ¢ = ¢available - 15 0 = 15° . This leads t o a mi ni mum r o c k
b ol t force of 2,300 tons in the direction 35 0 ab o ve t he hor izonta l.
When the distribution of normals is known, t h ere is a prob a -
bility of safety associat ed wi t h each factor of sa fety. Even t ho ugh
one might prefer to desi gn for a certai n va l ue of FS , he will prof i t
from computing the corresponding value o f PS. In the above example,
a probability of safety of 0.96 corresponds to FS = 1.7, while the
case with PS = 0. 99 agrees with FS = 2.15.
We h ave considered the plane of sliding to be cohesi on les s;
however this limitation can be overcome by cal cu l ating the va lue of
cohesive f o r ce ne ede d to reach the desired condition of stabili ty
(FS or PS). If the available friction angle is 30 0 , the re su l tant
of t he norma l for c e and friction on the plane can be i n clined a s much
Goodman- Geo log ical Eng. -17
244 Applications of Stereographic Projection

as 30 0 from the normal. If the factor of safety is to be 1.0 without


rock bolts, the frictional resistance will have to be augmented by
a cohesion of 8.0 tons per square meter, as shown in figure 6-24b.
To achieve a factor of safety of 1.7, the resultant of frictional
resistance and normal force will be inclined at 18.5 0 from the normal
whereupon the required cohesion is determined respectively as 11.1
tons per square meter (figure 6-24b). For a probability of safety
of 99%, the resultant must be inclined 15 0 from ~ , demanding a
p
cohesion of 12 tons per square meter.
It may be necessary to consider static water forces on one or
another of the faces of a potentially sliding block. What water
force on plane p (figure 6-24) will cause slip assuming an initial
factor of safety of 1.7? The water pressure acts in the direction
A

opposite to n . As shown in the right side of figure 6-24b, a


p
factor of safety of 1.7 implies that the resultant is inclined 18.5 0
A

from the direction of the outward normal (n ).* To produce slip,


+ P A
the force U shown in the direction opposite to n must cause the
resultant to be inclined 30
0
from ~.
p
The reqUi~ed water pressure
(lui/A) is 5 tons per square meter.
Similarly, an inertia force produced by a static acceleration
equal to K times the acceleration of gravity (g) will incline the
y
resultant to produce a factor of safety of 1.0. We can calculate
the value of K if we know the initial factor of safety. For example,
y
if the initial factor of safety is 1.7, slip occurs when the accel-
eration becomes equal to O.llg (i.e., K = 0.11). The critical
y ~
direction for the inertia force is shown by the vector Ky W in
figure 6-24b (the same direction as the minimum bolting force for
a factor of safety of 1.0).

ESTIMATE OF THE DISPLACEMENTS OF


A BLOCK UNDER A DYNAMIC IMPULSE

A dynamic load that produces an acceleration greater than K g


Y
given in the example above will precipitate slip but. the ground

* The "outward normal" is directed out of the block (into the


supporting plane).
Total displacement of th e ground below block
Xg = ~ Ag t~ + Ag to( tm - tal

-
Ground acceleratio n :Xg Tot al displacement of block
Base 1 2
di splacemen t Xb = "2 Ky g tm
Ag 1 - - -- - .....

Sli p between block and ground: llX = Xg-Xb


_ _ Yiel~ Cl£c~er:gti£n _ 1 1
.~ Kyg 2 2

-
~
aL...
de termi ned fro m
stereo net (see f ig .6 - 22 b)
!1X =-"2 Ag to + Ag to tm - "2 Ky 9 t m

Q)
u
u
a Relative moti on between block and support
t ime ceases at time t m
Kyg tm = A 9 to
tm = ~ or V
Ky Kyg
A 9 to: V and tm= _ V
_
Ky 9

time
Compare /).X to peak displacement in a shear test

Figure 6-25. App roximate


method for estimati ng the
signif icance of a dynamic load
for a potentially sl iding block;
after Henron et al (1971).
246 Applications of Stereographic Projection

motion may come and go before the block has travelled very far, as
discussed by Goodman and Seed (1966). Hendron et al (1971) proposed
a simple method of calculating the relative movement of a rock block
once the value of its yield acceleration K g has been determined.
y
The ground acceleration and its duration, or the ground acceleration
and the particle velocity, must be known. Suppose, as in figure
6-25, the ground acceleration rises instantaneously to a value of
Ag, remains constant for time to and then drops instantaneously to
zero. The block begins to yield when the ground acceleration reaches
Ky g, and acquires velocity Ky gt ' Relative displacement between the
O
block and the ground (slip) occurs until time tm when the block ac-
quires a velocity equal to the ground velocity V = Agt O;

V
tm K g
Y

The total displacement of the ground at time t is


m

At time t , the total displacement of the block is


m

x
b
= (1/2) Ky gt m2

The total distance of slip due to the ground motion is:

6.x

Substituting to = V/(Ag) and tm V / (K g), give s


y

6.x
~: (iy - i ) (10)

If the yield acceleration K g for slip in one direction is greater


y
than the yield acceleration for slip in the opposite direction, as
would be the case on an inclined plane, the block will slip uni-
directionally with each pulse coming to rest between cycles of
A pplications of Stereographic Projection 247

ac celeration. Equa t ion 10 is no t st rictly v ali d for an inclined p l ane


as the di r e cti on of re lative motion ca lculated is no t the same as the
down slop e d i r e c ti on of s l ip ; howe ver, i t c an be used to es t imate
the seve rit y of a p ar t i cu l a r impul s e on a block having a known yi e ld
accelerati on Ky g. If ~x ca l c ul a te d by equa t ion 10 is g r e a ter t han
the peak dis p l ace ment in a she ar test of th e block on t he surface,
the ground mot ion mi gh t creat e damage. Wi t h each incremen t of
displacement the e f fe ctive fri c ti on an gle, and the value of Ky f or
the next i ncrement , wo u ld b e r e duc e d a s demo nstrat ed experimentally
by Go odman and See d. We saw previousl y ( fi gure 6-24b) how to deter-
mine the yield acc ele r at i on ( Ky g) f o r unrestrained blocks on a single
plane. We wil l now e xte nd th i s t o sing l e tetrahedra l blocks resting
on two or mo r e p lanes.

SLIP OF T ET RAHEDRAL WEDGES

I n te rse c t i ng d is con ti n uity surfaces may liberat e tetrahedral


wedg es as in f igure 6 -26, a t t he top of a r ock cut. Two o f the four
faces of t h e we dge are free s u r f a ces. The wedge can slide in thre e
modes on eithe r of t he t wo suppo rt ing planes independently, and
o n b o th planes s i mu l tan e ous l y ; in th e l att er eas el s liding is parallel
to the line of i n t ersec t i on of t he two supporting planes . Condi-
tions of s l iding p a ral lel t o t he l ine of int e rsectio n were discuss e d

Figure 6-26. Limiting orientation of


the resultant at limiting equilibrium
when tending to slide on t wo planes
simu lta neously.

Intersection mode slip


Result fo rce between wedge and supporting
planes IS R
Oire,ftion of shear resistance In each plane
IS II'2

At limi ting equi li brium

R = R, +R2
= (N,n, + T, I'1r-( N2n
1 + T2 I~ 'l.)
248 Applications of Stereographic Projection

by John (1968) using the stereographic projection, and previously


by Wittke (1965) using vector analysis, and Londe (1965) using other
graphical methods. The mechanics of sliding on either plane alone
is identical to the examples discussed previously except that there
now is a kinematic constraint -- sliding on a single plane is possible
only within a restricted set of directions in the plane, as there can
be no net displacement towards the adjacent face. If one plane
provides the entire reaction to the resultant force initiating sliding,
the resultant can be decomposed uniquely into components normal and
parallel to that plane. Since the block cannot move into the adjacent
plane, the direction of the shear force must lie between the inter-
'"
section 112 and its opposite in the half plane containing the outward
normal. Thus, on a lower hemisphere stereographic projection, con-
struct the normal to the plane in question and the line of inter-
'"
section 112 between the two planes (figure 6-27); the great circle

part of friction circle


for plane 2

slides on
plane 2

The resultant force ot


the wedge on its supports
is safe if its direction is
inside the ruled area

L.H

Figure 6-27. Generalized friction circle for sliding of a wedge with two faces in contact.
Applica tions of Stereographic Projection 249

",,...--.....,,,
112 n 1 bounds t he re g i o n of kinematically poss ib le sl ip on plane 1.
A~
Similar ly th e gre at circ le 112 n 2 bounds the k i ne matically me an ingf ul
portion of the f riction circle for slip on pl an e 2 alone (fi gu re
6-27) , Between these great circ les the r e is a regi on in whi ch slip
is pos sible a long th e l i ne of int e rs e c t i on; in t hi s case, the direc-
t ion of in ci pien t s li d in g is parall el to the l in e of int e rs e c t ion
an d t he re fore the sum of the shea r f orces on each plane , at the l imi t
of slip, is p aral lel to '"112'* In a dd ition there are fo rces paral lel
to the two normals. The r e sultant f orce of the blo ck on its s up p orts,
in t he case of sliding d own the l ine of i n terse c tion , must t he ref ore
'" A A
b e composed of c omponent s paral lel to n 1 , I , a nd n and it will
12 2
acc ordingly pl o t i n the spher ic al t riang l e fo rmed by
1
, 1 , an d
12
n
n . If it is possible to slide up the line of i nt e rs ecti o n , the
2
re su ltant f or c e will h ave to lie in th e s phe rical tr i a ngle b e tween
n1 , -1 12 a nd ~ 2 (f igure 6 -2 7 ) .
At the l imit of e qu ilib ri um in the i nt e rs e cti o n mode, ful l
friction has been mob i lized in both pl anes simul t an e ously. The
re action in plane 1 is ma de up of a n ormal fo rce an d a s hear force
A 0 '"
paralle l to 112 a nd is ¢1 from n 1 . Th e r efo re t he r e a c tion in plane
1 lie s at the i ntersection of the f ri ction circle for plane 1 with
A ~'"
the great cir c le 112 n 1 · Si milar ly the r e action in plane 2 i s made
A

up of a no r ma l forc e, and a she ar fo r c e para lle l to 112 an d is


inc lined ¢2 ° f r om th e norma l to pl a ne 2; therefore t h e r e action in
plane 2 wil l be fo un d at the int e rs ec tion of the fri c t i o n circle for
",...--....."
plane 2 a nd t he great circ le 1 12 n2' Si nce the res u ltant fo rce i s
c ompris e d u ni q uely of the re ac tions in pl ane 1 and plane 2 it must
l ie along th e gre a t circ le common to t he se two react ions , as s hown
in fi gu re 6-27 . Gi ven the fri cti on ang les on p lan e 1 and p lane 2 ,
therefore, it i s p ossible to construct a ge n eral i ze d safe zone in-
cludi ng al l possibl e modes of t ranslat io n. Once construct e d, this
safe zone is u s ed in t he same fashion as il lus t r a t ed previously for
the simple f ric t ion c irc le. One e xamine s the or i en tat io n of t he
resul tant of app lie d f orces wit h re spe ct to t he s af e z on e. According

* Initial s h ear s tresses th at may have exist ed must change to


the limi ti ng v a lues an d d irec tio n s be fore slip and theref or e
they are of litt l e i ntere s t fo r the li mit eq ui l i brium analysis.
250 Applications of Stereographic Projection

to the position of the resultant not only the degree of safety, but
the mode of potential slip is indicated.

Meaning of the Factor of Safety in the I ntersection Mode

As discussed by Londe (1965), there is an ambiguity attached to


the discussion of the factor of safety when an intersection mode
slide is in question. With respect to sliding on a single plane,
A

the factor of safety implied by a particular orientation R of the


resultant force, inside the safe zone, was found by shrinking the
friction circle until its circumference passed through R. However,
for intersection mode sliding, there are an infinite number of ways
that the safe zone can be reduced to bring its perimeter exactly
through R. In the illustrative problem, presented in figure 6-28a,
we are presented with the normals to two planes and two friction an-

(a) North

o
111
-8

Figure 6-28. (a) Example of


intersection mode analysis.
(b) infinite combination of
friction angles that will pro-
'" opposite to
rock bolt
direction for
minimum force
required.
vide limiting equilibrium
cond itons. (c) I nclusion of
cohesion and rock bolt forces.
Applications of Stereographic Projection 251'

(b) -N
Factor of saf et y on plane I
9- I"- <.D N N m 0 <;;t- o
II (J)
~ ~ ~ LO ~ N 0
--- - - - -
25 1.0
E I
::l
~

..0
=-= 20
V
I
FS 1 0 FS 2 0 1.85
1.12

1. 28 C\.l
::J
Q)
0-
Q) I 1,48
c
0
~ I 0.
E I c
15 1.74 0
10- ---------- >.
0 ~
'+-

"'0
2 .10 2
0
Q) (j)
1-
- 10
::J 2 .6 4 '0
0- 1-
~ 3 .57 .2
(,)
N
9- a
--
LL

0
5
a>
::J
0
>
0
0 5 10 15 20 25 30 35 ( = CPI)
Value of tpl required for limit equilibrium

(c)

o
o o
o o
II

,(r <Pava i I able= 10° on ea c h plane

I(TI + C + T2 + C2 ) 1/ ITI =
1 1.3

B=1 4
252 Applications of Stereographic Projection

/\

gles. The orientation of the resultant given, R, lies well within


the safe zone and therefore the wedge is stable. The dashed lines
are great circles obtained by reducing ¢1 and ¢2 by different amounts
to shrink the safe zone through point R. The values of ¢1 and ¢2
required for limiting equilibrium (¢ . d) are then plotted in
requlre
figure 6-28b. One may adopt a convention that the "factor of
safety" is the value which produces the same factor of safety on
both planes simultaneously, which in this case is 1.85. However,
there is no theoretical justification for such a definition . It
is more meaningful to use the curve of ¢1 ,requlre
. d versus ¢2 ,requlre
. d
as the expression of safety and, in particular of the sensitivity
of the safety of the wedge towards the change in either of the fric-
tion angles.
A probabilistic interpretation of safety as in the example with
a single block on a single plane, is more difficult in a case of
intersection sliding. This is because a given dispersion of normals
about planes 1 and 2 produces a greater dispersion of their inter-
sections (figure 6-29a). This problem was discussed by Ramsay (1967).
It will be seen in figure 6-29b that as the angle between the normals
becomes small, the dispersion in the position of the line of inter-
section becomes very large. We can shrink the friction circle on
each plane corresponding to the degree of probability required; but
it is more difficult to draw the required great circles connecting
the normals with the line of intersection, for as the normals wander
the line of intersection wanders even more.
The generalized friction circle, i.e. the safe zone is used for
a tetrahedral wedge just as for a single block on a plane, as illus-
trated in figure 6-28a. For example, if the available friction angle
0
is known to be 10 on each plane, we can ask what cohesive forces
are required to insure a factor of safety of 1.5 on each plane. The
bounds of the safe region in the intersection mode corresponding to
o /\
¢l equal ¢2 equal 10 are shown in figure 6-28a. In thi~ case R is
no longer safe and it will be required either to rotate R into the
safe zone by addition of force or to increase the size of the safe
zone by assuming a cohesion. A solution to the latter problem was
presented by John (1970). By using the construction of figure 6-14,
A pplications of Stereograph ic Projec tion 253

Figure 6-29. (a) The dispersion of li nes NORTH (a)


of intersection of planes, each of which
is dispersed 10°. (b) Dispersion of
intersections as a function of 11 and 1/1.

(b)
80

N
.....-
'0 60
c
9(/)
'-
(l)
0...
(/) 40
a
><
0
E
(l)
20
i
----!
0
0 20 40 60 80
T} - Angle betwee n no r mals 11 " n2 { Ram say - 1967, p. 14 1
254 Applications of Stereographic Projection

decompose R into three components -- one in the direction of sliding,


A

1 12 , and the others in the directions of the normals n and n . The


1 2
resultant normal force N is found from the orthographic projection
of
R on
plane n 1 n 2 .
A~A A
(Since 112 is normal to this plane, the inter-
A~A ~A
section of great circle 112 R on plane n n is the same as the
1 2
orthographic projection of R
on plane n~2') Figure 6-28c shows
the polygon of forces for this decomposition. Once the normal has N
been determined, it is further decomposed into normal forces in each
-+ -+ 0 0
plane, N1 and N2 , making use of the angles 41 and 30 measured in
plane n~2' Figure 6-28c shows how the cohesive forces are added to
the force polygon to find effective overall friction angles on each
plane. If the friction angles mobilized in these planes are both
10 degrees and the cohesive forces on planes 1 and 2 are respectively
14 and 9, the ratio of resisting to driving forces is 1.3.
Rock bolts can be used to rotate the resultant force into the
safe zone so that the mobilized friction does not exceed 10 0 in
either plane. The minimum rotation of force required to produce the
desired safety, without cohesion, is 8 0 (to point 8, figure 6-28a).
The direction for bolts to effect this rotation with minimum force
is B, (in the direction opposite to
~
-B of figure 6-28a), that is 90
0

from 8 along the great circle R 8. As shown in figure 6-28c, the


required bolt force is 14. B rises 33 0 to the N52 E.
0

Problems with intersection mode sl~ding in argillite were met


by the Corps of Engineers in construction of Libby dam in Montana.
Intersections of bedding and jointing plunged 34-38 degrees into
excavations on the left abutment (figure 6-30a). Figure 6-30b shows
the supporting planes of a wedge approximately 600 feet long after
slide debris had been completely cleaned out (see also figures 2-8d
and 2-8e). One potential intersection slide involving the upper part
of the left supporting plane (bedding) was stabilized with an installa-
tion of approximately eighty 200-ton tendons, anchored 65-145 feet
deep; these are visible in the upper left of figure 6-30b and in
figure 6-30c. Stereographic projection of the wedge shows immediately
the potentially dangerous situation, since with only gravity loading,
0
a friction angle of approximately 30 on each plane is required to
maintain the wedge.
Applications of Stereograp hlc Projection 255

(a) (b) (c)

Figure 6-30. (a) A dayJighting intersection between a bedding fault and jointing
at Libby Dam, Montana. Photo by Dennis Lache l; (courtesy of the Corps of
Engineers). (b) A portion of the left abutment slide at Libby Dam after removal
of the slide debri s. T he bedding, left. strikes N 26 Wand dips 43 -5 to the west.
The joi nt su rf ace (right) st ikes N 74 E and dips 49 to the NW. Figure 6-30a was
t aken just t o the left of this view. Compare with figure 2-8d and e. Photo by
Dennis Lachel ; (courtesy of the Corps of Engineers). (c) View from the top of
the wedge of 6-30b showing the relatio nship of the slide t o the dam and reservoir.
Photo by Dennis Lachel (courtesy of the Corps of Engineers) .

SLIDIN G OF TETRAH EDRAL WEDGES WITH ONLY ONE FREE SURFACE

Lan d e (196 5 ) , a nd Lande e t al (1970) discussed t his problem in


the context of t he ab ut me nt st abi li t y calcu l ation s for an arch dam.
Indeed it i s bel i eved t h a t t he movemen t of a block r esembling tha t
of figu re 6-31 , permi t te d t h e fa i lure of Malpasse t Dam (figure 1 - 3).
On e must a do pt a s i gn con ven t i o n for the directi ons of th e li n e
of intersect i on and t he n orma l f orces . A l in e of i ntersection wil l
b e c onsi dered positive if it poin t s int o the f re e space , i.e . if
it dayli g ht s . It may b e in t h e upper or i n th e l ower hemi s phere
according to t h e o r ie n tati on o f t he we dge in space . Figure 6-3 1
shows a tetrahedral b l ock wit h one fre e s ur f a c e and indicates th e
posit ive di re ctions of the lines of intersection and th e normals;
256 Applications of Stereographic Projection

sketch shows 1he positive


directions of intersections
Figure 6-31. Kinematics of and normals (sequence
of indices is not
a wedge with three planes important)
in contact.

as before, a normal is positive if it points out of the block into


the supports. Now there are six translational modes -- three modes
of sliding on a single plane, and three modes of sliding along lines
of intersection. A mode will be named according to the face which
opens to permit the translational movement, according to the ter-
minology of Londe (1965)*. Mode 1 signifies that face 1 opens, and
all reaction to sliding is contained on faces 2 and 3. In general,
mode i signifies opening from face i with sliding on planes j and k
r-
in the direction of the intersection I jk . A mode of sliding on a
single plane requires opening from the other two and therefore will
be called mode ij with sliding on plane k; for example mode 1,2
necessitates sliding on plane 3. Each mode of sliding is confined
to a particular spherical triangle of the reference sphere, as
indicated in Table 6-1. For the mode of sliding in one plane, the
resultant force will decompose into components in the directions
of the normal to the plane of sliding and the lines of intersection

* In the work presented by Londe (1965) the sign convention for


normals and lines of intersections was opposite to that pre-
sented here. However the result is the same, for rather than
the orientations of resultant forces applied to the block, Londe
mapped the orientations of the reaction force of the supports
on the wedge.
Applications of Stereographic Projection 257

TABLE 6-1

Mode Slides on: Opens from --


R is inside

0 none none
A

nl '"
n
A

2 n3
A

I 2,3 I n 1 n3
2 23

2 1,3 2 i\ 113 A3

3 1,2 3 AI 112 rl2

1,2 3 1,2 113 A3 123


2,3 I 2, 3 II2 AI 1
13

1,3 2 1,3 112 rl2 123


1,2,3 none all 112 123 1
13

along either of it s e dges . Th i s is bec aus e of t he k inemati c re-


q u i rement t h a t the d irection of mot io n lies be twee n t h e t wo edges
of t he sliding face. For exampl e, for slidi n g on pl an e i t he
re sul tant force mus t be in t he sp h erica l tr i ang l e I ij , n ,
i
Iik. The
three such mo de s a re i d en ti fi e d in Table 6 - 1 . For sliding in t wo
plan es si multan eously , as previ ous ly d i scus se d for t h e two p lan e
wedge, the result ant fo r ce will decomp ose in to a norma l f o rce on
e ach pl ane an d a she a r force paralle l to the line of i n tersect io n.
There fore for mode k wi t h sli di n g on planes j and i , R mus t l ie
insi de t h e spheri ca l t ri a ng le for med by ~ i' I ij , TI j . The t h ree
such mode s are pres en te d in Table 6- 1 Two sphe r ical t riangle s
remain un l abel l ed: t h e spherical t r ian g l e ~1 TI2 n3 wh ich is com-
p le te l y safe (mode 0) ; and t h e sphe r ical t r i an gle I 12J 12 3 , 11 3 in
whi ch the we dge must li ft out ( mo de 1, 2, 3 ).
F igure 6-32 p r e sen ts an e x ampl e . A rock we dge f ormed b y t hre e
intersecting pl a n es lies in th e a but ment of a n arch dam; its we i gh t
a nd the l oading of t h e d am are given, a nd it is desired t o calc ulate
the wa t er fo rce on f a ce 1 wh i ch c ou ld initi a te f ai lure of t he we dg e.
The ap plication of Table 6-1 to this problem is facil i tate d by
re cogn izing t he var ious norma ls an d l i ne s of int ersection s (Table
6-2 ) all o f which c an b e ob tai n e d f rom a s ubsi d iary st e r eogr a ph ic
258 Applications of Stereographic Projection

TABLE 6-2

Line Bearing Plunge or Rise Hemisphere

nl west 0° Both

n2 North 20° up upper

n3 - vertica I Lower

112 South 70° up upper

123 East 0° Both

113 South 0° Both

projection, (or in this simple case by inspection). Figure 6-32b pre-


sents the lower hemisphere and figure 6-32c the upper hemisphere

(a)

7P®
Q)ho ,@ horizontal

Face Strike Dip Friction angle Figure 6-32. (a) Example problem of a wedge
I North vertical 30° supported on three faces. (b) Lower hemi-
2 East 70° 50° sphere portion of the generalized friction
3 - HOrizontal 20° circle (the safe zone) for the example problem.
Applications of Stereographic Projection 259

portion of the reference sphe r e . The sph e rical t riangles of Tabl e


6-1 are formed by constructing the requi re d g reat ci rcl es. Given th e
friction angles on each of the p lan es , we c a n c onst r uct the gener aliz ed
o
frictio n c i r c l e for t he we dge in q uesti on . Suppos e ¢l = 30 ,
¢2 = 50° an d ¢3 = 20° . Small ci rc les of the r e q ui re d radii are co n-
structed in the proper sphe rica l t r iang les. That is , c onstruct : a
s mal l c ircle of radius 2 0
0
in the triang le a
i 1 3 , 123 , 3 ; a small
0
circle of rad i us 50 about 02 in spheri ca l tri a n g l e 112 , ll2' 123 ;
about ~1 in t he s ph e rica l triangl e
0
:n d aAsma! l c irc le of r adius 30
1 n , 1 . Draw g re at circles betwe en the intersections of the s e
13 1 12
small ci rc l es wi th the mode boun da r ie s as sh own . Th e on ly diffi-
culty arises in connection with the g r e at c i rc l e from poin t b'" in

NORT H

(b)

1----------- safe ----+--:;,------------....,

nl~____________~______+-__~

Goodman-Geol ogi cal Eng -18


260 Applications of Stereographic Projection

(c) NORTH

U.H.
-w

W= 10,000 tons

water force on face


to produce slip is
Figure 6-32. (c) Upper hemisphere portion U, = 5300 tons
of the general ized friction circle (the safe
zone) for the example problem. (d) Force
slip will be in mode
polygon for the solution to the example
problem. (sliding on 123 )
Applications of S tereographic Projection 261

the upper hemisphere to point a in the lower hemisphere representi n g


the limits of the s a fe zone in mode 1. In order to construct this
great circ l e plot -a (the opposite to ~) in the upper hemisphere p ro-
~

jection and dr aw the great circle -a b; this great circle, in the mode
....
1 region, intersects the primitive circle at point d, wh ich can b e
transferred to the lower hemisphere. The required great ci r cle
construction is completed by connecting d and a.
The completed c o nstruction shows that the orientation of the
resu l t ant force, R, is well within the safe zone. If t he wat er
pressur e is raised on face 1, a water force on the we dge wi ll poin t
in the direction of the opposite to ~l' (-~l = 123 in t h i s particu la r
problem). Along the great circle a:'ft a 14 0 rotation is requ i red to
bring the resultant force to the l i mit of stability. R, found by
addin ~ the weight force (in direction ~~) and :he thr ust of t he arch
d a m (Q) i n the di re cti on given, lies 81 fr om U1 . The polygon of
forces o f fi g u r e 6-32d shows that the required water force to initi-
ate sl ip i s 5,300 t o n s and failure will be by mode 1. Ro t at ions
have not been considered; in any p r a ctical example one cou ld u s e
the methods discussed p r eviously for detailed analysis of rot a t i ons .

SLIDES COMPOSED OF TWO BLOCKS

In general a rock mass consists of numerous mati n g blocks with


lit t le pore s p ace. The movement of one block creates the f r e e d om
for its neighbors to move and therefore general pr ob l e ms of rock
stability require a n appreciation of the interactio n s of multiple
blocks. We will o n ly begin to explore t h is question he re , using
the stereographic projection. Figure 6-33 shows a slide consi st i n g
of two free b lo c ks, one acting as a passive supporting wedge and the
other acting as a n active loading wedge. Sliding of the passive
we dg e is promoted by the weight of t h e a c t ive wedge, which c an no t
be sustained by frict i o n on its base alone. Thus, there is a load
transfer from passive to active we dge. Figure 6 -3 3 is a secti o n
through the two blocks in the direction of sliding. It need not be
parallel to the dip vec t o r i n ei t her plane as will be disc ussed in
the later e xample. A potential two block problem can oc c u r only if
t h e sliding surface of the active wedge daylights into t h e s p a ce
262 Applications of Stereographic Projection

Figure 6-33. Two block slide.

created when the passive wedge moves. Figure 6-34 shows a simple
analysis for the mechanics of this case. The principal idea is to
separate the blocks along their common plane (plane 3 in this example)
which is considered as a free surface in each analysis of each block
separately. As opposed to the case in soil mechanics where the
division between the active and passive wedge has to be determined
by iterations, structural geological information determines the
direction of the interface between blocks. In contrast with soil
mechanics, there should be no debate about the proper angle of
friction for the inclination of the load transferred along the con-
necting plane, since a very small displacement is enough to mobilize
Applications of Stereographic Projection 263

ACTIVE B LO CK (A)

PASSIVE
BLOCK (P) cross secti on of blo cks
in direction of virtual
motion

Analysis:
I. Analyze acti ve block wit h
plane 3 as a f ree f ace .
Find Fp
requ ired.
2. Ana lyze passive block with
pl ane 3 as a f re e f ace ,
and with addi t iona l loa d
- Fp .
3. Safe if re sulta nt on passive
bloc k is in safe zone .

PASSIVE B LOCK ACTIVE BLO CK

LH LH

Figure 6-34. Analysis of a two d imensional t wo block case.


264 Applications of Stereographic Projection

NORTH

Example - Two Block Slide


Given
WA = 10,000 tons
30 0 W p " 5000 tons
'\. - safe R1 = 35 0 down to N50E
- \;R - zone
r--=~-- 1 for n2 = 80 down to N 50 E
0

-\-:",~~---- active np,3 = -nA,3 = 50 up to N 50 E


block 0 0
1'1 = 1'2 = 30 ; 1'3 =15

IS the system of blocks safe?

No.

Figure 6-35. Example of a two block analysis; two dimensional case.

the full available friction. Therefore, we will take ¢3 as the


angle of inclination of the load transferred between blocks, as shown
in figure 6-34.
The active block is analyzed in the right half of the stereo-
graphic projection in figure 6-34. It is a simple friction circle
case with sliding on a single plane. The active wedge is unstable
under its own weight; we require a transfer of supporting force from
the passive wedge, which is inclined ¢3 from np, 3 -- the direction
of the normal pointed into the active wedge. The direction of the
transferred load (fp) is determined from a force polygon to rotate
the resultant through the required a degrees to bring it into the
A pplica tions of Stereographic Projec tion 265

sa fe z o ne . Pl ot the opp osite of f i n t h e st e re og r ap h i c projec ti on


p '"
for the p as sive bl ock (its d irection is - f ). Addi ng t he weight of
p
th e passi ve bl o c k t o -F produces a r esul tan t R2 i nc li n e d wi th respect
7.. A P
to wp ' If R2 is i n t he safe zone f o r t h e pas s i v e we dg e , then both
ac t ive and passi ve b l ock s are safe. We will n o t t ry t o balance
f a c t o r s of safety b e tween wedges, but r a th er proceed t o a n umerica l
e x ample (fig ure 6- 35). Th e active we dg e wei g h s 1 0,000 tons , the
pas si ve wedge wei ghs 5, 000 tons, a nd t h e dire c t i ons of the no r mals
to p l anes 1, 2 a n d 3, are given as wel l a s the fr i ction angles. Is
t h e system o f blo ck s safe? Bo t h t h e act ive and p ass i ve we dge con-
stru c t ions h a ve be e n s up e r impos e d on a sin gl e s te reogram for t his
pr o bl e m. Th e weig h t of the active we dge i s incli n e d 25 0 f rom the
safe zone for t h e a ctive block an d t h e refore t he force Fp s hown
in the f o rc e po l yg o n is req u i red. Th e a d diti on of Wand
p -F p pro -
0
duces a 2 8 rotati o n of the wei g h t fo rce i n t he p a ss i v e b lock, which
p r oduce s a r e su ltant outside of the safe zone for t he passive block .
Th erefore t h e s ys te m of two b l o c ks is unsafe.
The a b ove e xample is an il l ustr ati o n of the p a r t icu l ar case in
wh ich ea c h blo ck s l i des on a si ngl e plane ( mo de i,j ). I f e ither , or
bo t h block s a r e in volve d in an in ters ection type s l ide , t he prob lem
is more di f f icult. Fi gu re 6 -36a illustrates such a c ase . The active
blo c k formed b y p lanes 1, 2 and 3 tends to sli de down i ts li ne of
intersection. Den o t e by q the poin t where the l i ne of i n t ersect i on,
/'-

I12 of the acti ve block meets the face o f the pass i v e b l ock . If q
is inside the face of t h is block, then movement of the passive b lock
will al l ow da y light ing of the active wedge, a nece ssary condition
for t h e t wo bl o ck slide. If t h e pas sive wedge sl ide s on a s i n gle
plane , the d i rect i on of relative s l i p on plane 3 wi ll have to be
determine d b y ite rati o n. If on the othe r h an d, b o t h t he active
an d p a ss i ve wedges move a l o ng t he ir li n e o f in te rs e c t i o n s , there is
o n ly one possib le di rection for the r el a ti ve sl i p in plane 3, a s wil l
b e shown.
Si n c e at first t h e critic a l mo de s are n ot known, an i terat ive
s o luti on will b e i l l u s i rated. In f i g u re 6- 36b , t he active wedge i s
diagramme d . We will s up p ose we do n o t kn o w t h e di r e ction of slip
on pl a ne 3 an d will determin e a n umb e r o f solut io n s f o r diff e rent
266 Applications of Stereographic Projection

Active wedge 1,2,3

passive wedge 3,4, 5

Analysis: Consider plane 3 a free face and determine direction


of slip along it as shown in figures 36 b, c.

A
For intersection sliding of active wedge, 112 must
(a) intersect face pqr of passive wedge.

Figure 6-36. (a) A two wedge slide. (b) Analysis of the active wedge. (c) Analysis of
the passive wedge. (d) Force polygons for analysis of the two wedge case. (e) Construc-
tion for compatible velocities of neighboring blocks, each of which slides on its line
of intersection.

possibilities. First construct the safe zone for the active wedge;
then around the normal n 3' construct a small circle of radius
equal to ~3' Points a, E: c, a, e and f are various possibilities
for the direction of the load transference between the active and
the passive wedge. Corresponding to each we can read the required
rotation a of the weight force to produce a resultant in the safe
zone of the active wedge, as tabled in figure 6-36b. Suppose ¢
App l ications of Stereographic Projec tion 267

NORTH
A

Locus of fp

ACTIVE WEDGE

Orientation of Angle Ro tation of W


A
passive Force between req uired
(fp ) fro and WA ( a)

a 95 16°
b 110 15 .5°
c 114 16°
d 114 16°
e 110 17 °
f 97 ISo

cp=20 0 0n all planes

0
equals 20 o n all plane s ; what an gl e of f r icti on is necessary for
stab i li ty in the pas si ve wedge ?
Sin c e the ang le b e t ween the direc-
"
tio n of fp and the weight fo rce of t he active wedge i s known f or
each as sumption of di rection fp , a f o rce p ol y g o n c an be cons tr ucted,
a s shown in fi gure 6 -36d , defini n g t he magn it ude of
" A A
FP corresponding
t o each a ss umption for f (a thr ough f ) . Tr an s fe r t he opposites to
p
p o i nts a th ro ugh f onto the st e reogr a p h ic p rojection f or the passive
we dg e ( fi gure 6 -36c ) as we ll as t he opp o s i te to
~ p,
3 (point
a,
3).n n
Read the a ng l e t o Wp f r om each point jus t p l otted and construct a
s er i e s of fo rce p olygon s f o r e a c h point to determine t he rotation
of W
p
p r o duce d by each assump t ion of s lip di r ect ion in plane 3.
This h as been d one ( fi gure 36d ) for two a s sumptions o f the weight
268 Applications of Stereographic Projection

(c) NORTH

PASSIVE WEDGE

Wp = WA and Wp = WA I2

WP=WA 1.
W =W 12 (~sultant
p:c\ ) \

'y--Plane 3

"~o \
"- '-. A \ \Direction 0 slip in 3
~~5 ,~ [I~ Aplane 3]
--- 29 0
- - .----.------,-----.-------

A
fp
Angle from
Af A
IFp I Inclination
of resultant
- p to wp
wp= WA wp = W,/2

a 85° 0.28WA 15.0 27 .0

b 70° 0.27WA 13.5 23.0

c 66° 0.28 WA 13.0 22.0

d 66° 0.28WA 13.0 22.0


Figure 6-36. (continued)
e 70° 0 .295WA 14.0 24.5

83° 0.32WA 17.0 29.5

of the passive wedge: W = W , and W = (1/2)W .


pap a
In a more general problem, other forces might be included, e.g.
the thrust of a dam, or water pressure forces on any given face. It
is seen that the resultant is unstable and the passive wedge must
fail by sliding in the direction of its line of intersection. The
direction of slip on plane 3 is therefore determinable as follows
A~A
(figure 6-36c): Construct the great circle I12 I 45 , and mar~ts
intersection, m, with plane 3. Then construct great circle m n a, 3'
Applications of S tereographic Projection 269

(d) (e)

Figure 6-36. (continued)

The i nt ers e ct ion of the l a tter with the f ri ction circle abo u t n
A a, 3
def i nes poi n t n wh ich is the direction of the resul tan t load t r ansfer
fr om the active to t h e passive wedge. (The direction ~ correspon ds
to the d irecti on of slip in plane 3.) The wedge is un safe. An
e xp a nsion o f the fr ict ion circle about ;i 4 to 37 degre e s will p r o duce
limiti n g eq u i librium.
Fi gure 6- 36 e shows the vector triang le, used t o construc t t he
270 Applications of Stereographic Projection

vector subtraction to determine point m. The relative slip in plane


3 was obtained by the difference of vectors pointed in the direction
of slip on each wedge, since it is the relative movement between the
wedges which provokes the load transference. The orientation of
A A

the vector difference 145 - 112 is known to be in plane 3 since the


blocks maintain contact during sliding (assuming there is no rota-
~ion). The relative magnitudes of the vectors parallel to 112 and
145 indicated by the vector triangle, figure 6-36e, indicates the
relative velocity of each wedge down its line of intersection.
Observation of the displacement of one wedge, therefore, can be used
to infer the compatible movement of its neighbor. In this instance
the active wedge moves 0.49 as far as the passive wedge in each
increment of slip.

THE STATE OF STRESS IN ROCKS

Rocks differ from most other materials in that they may already
possess significant stresses before additional loadings or un loadings
are constructed. Excavation of surface or underground space leads
to stress rearrangement with concentration or spreading of the lines
of force, which, to the extent that the rock properties vary with
stress, render the rock mass non-homogeneous. If stress concentra-
tions cause extreme stress differences, the rock may break, sometimes
explosively as in deep mines in South Africa and Canada. On the
other hand, if stress reductions leave blocks of rock almost free
from their neighbors, they may fallout, as in the roof of a tunnel
at shallow depth and in the exposed corners of excavated rock walls
in complex underground openings like underground power plants (see
figure 1-5b). Since properties of joints are highly stress dependent,
a jointed rock mass with a non-homogeneous stress field will change
in character from point to point. In such rocks, we are particularly
concerned with the real possibility that a state of stress may imply
sliding or opening of individual planes. Whether or not such move-
ments will provoke rock falls depends upon the direction of the
sliding or opening tendency with respect to the kinematic freedoms
of the rock mass. Given a state of stress, we will show how to
calculate the normal and shear stresses local to a given discontin-
Applications of Stereographic Projection 271

uity, and how to depict their directions with re s p e ct to the p l a ne.


The stereographic projection is useful here, particul arly as r egards
the directions of shear stresses.

Stress

The state of stress at a point in a solid is de s cri b ed by the


set of three vectors representing the force per unit of ar ea tr a ns-
mitted across th ree perpendicular faces radiating f r om the point in
question. Figure 6-37a shows the naming convention for these vec t ors
a n d t h eir components.
"-
Choosing a xes arbitrarily in orthogonal direct i on s x, y, an d z ,
t h e plane perpendicular to ~ is called the x plane and th e vector of
+ 7
f o rc e per u n it are a across it is P. It is usual to represent P
x x by
t h e following th r ee components: one normal to the x plan e, a ;
x
one in the x p lane pa r allel to y , T ; and one in the x p l an e p a rallel
xy
to z, TXZ The state of stress is thus represented by nine c ompon-
ents; but consideration of moment equilibrium about each axis in t urn
leads to three resu lts of the form T T and t h us the comp l et e
yx xy
state of stres s at a point is designated by six indepen dent com-
ponents.*

aX T
yx T
zx
{a } TaT (11 )
xyz yx y yz
T T (J
ZX yz z

It the state of s t ress is kn own with respe c t t o axes x , y, and z, t he


vector of force per unit area may be computed for a plane in a n y other
orientation, e.g. wi th normal parallel to ~ , and contain i n g p erp en-
dic ul ar l i n e s y' and z'. It will be necess ary to determine t h e
angles: f rom ~, to x, y, A

and z; from y' to x, y, and z; a nd f rom


z' to x, y, and z. The table of dire ct ion angles is most easily
de termin e d usi n g stereographic projection, as in figure 6 - 3 7b. De not e
the cosi nes o f the first three angles as 1 , m , and n , of t h e
1 1 1

* The brackets { } denote a column matrix; arr a ys are enclos e d b y ( ) .


Figure 6-37. (a) Components of the state of (a)
stress at a point. (b) Stereographic projection
z
for direction angles. (c) Positive normal and
shear stresses on the x' plane.

(b) NORTH
x' plane
area A

,-
,-

/
./ ""
x' plane ',\,/
/
/
/
/
/
/
/
/
/
/
",y./ I
/
I
I "y
I
I
/
I
I
I
I
I
I
I
I
I
I
I
,
1

I x y
x' 65° 42° 60° N

"x y' 104° 116° 3d' v,0 x' plane

Upper Hemisphere z' 30° 120° 90° 60°


Appl ications of Stereographic Projection 273

Xl
(c) z

x', y~ 1.' {
components
of Px'

x' plane

second three a ngle s as 1 , m , and n , and of t he third se t as 13 1


2 2 2
m , a nd n 3' The x , y , a n d z component s of t h e force per uni t area
3
a c ros s the x' plane ( P x I) are :

(12 )

The e qu i l i b r ium of the wedge shown in f igu re 6-37a demands that t he


for c e pe r unit are a of the x' pl an e transmit ted from the n e ighboring
mater i a l i nto t h e wedge a c r oss t he x' p l a n e is the opposite of Px I

the ve ct or re s ult a n t of t h e components cal culat ed by equat ion (12).


It is use fu l to decompos e Px I i n t o c omponen ts in th e direction x '
n orma l t o t he x ' p l ane, and in directions y' and Z l c hosen conven-
ientl y in the x' plan e; thi s def in es st r ess components on t he x'
plane:

(1 3)
274 Applications of Stereographic Projection

where

(T)

We may take positive ax' to signify compressive stress. Then if ax'


is directed in the positive coordinate direction, the direction
of positive x', T ,
x y
, and T ,
x z
, are positive in the positive y' and
z' directions respectively.
As an example, given stresses referred to x, y, z axes (figure
6-37b )

first we find

.422 .743
.500)
(T) -.242 -.438 .866
(
.866 -.500 o

By equation (13)

(6.2, -1.8, 2.1)

A complete stress transformation to new coordinates x', y', z'


requires repetition of this procedure for each of the three ortho-
gonal coordinate planes. The result is

( a)x' y' z' (14)

Further Use of Stereographic Projection


The above procedure completely solves the stress transforma-
tion problem but gives a mathematical rather than physical result.
Applications of Stereographfc Projection 275

The s tereograph i c projection shows how the stresses are ori e nted on
the plane o f i nterest .
Consider the x, plane of figure 6-37b, whose direction cosines
were previously obta i ned . The x , y, and z components of the traction
~ can b e c omputed fro m e quat i on (12)
I gi v ing f or the example
x
computation

( 4 . 96, 4.39 , 1.49)

Using th e methods previously considered for addi tion of three force


c omponent s (figure 6- 1 3) we can obtain t he magnitude and direction
of the resul t ant P ,. On the ste r eographic proje c tion (figure 6-38) ,
~
A is the orientation o f
x ± *
+ P , ) and B is the or i en t ation of
(p,
A

(P x r x + Px ,z) ; th e y can be-1 p l ottedx using


± y x x
the angles 0
xy
= t an
-1

(P jP ! ) and a
I = tan (P IP . ). Then the direction of P r
I
X y X X xz x z x'X ,('""-A ~ x
is ~ located a t the i nt e r s ectio n o f gre a t circles A z and B ~
I

whil: the magnitude of Px I is Ipx " I = (P I


x x
2 + P I
x y
2 + P , 2)t
x z
In the example pre viously considered, 0 xy = 41.5° , 0 xz = 17.5° and
IP x ' / = 6.8.
Finally , the orientation of the maximum shear s tress in the x '
plane is i x r ,max found as the nearest intersection of the x ' p l an e
and the p l an e c ommon to X' and Px f' As shown in f i gure 6-38,
0
T is in t h e l ower hemisphere pi tch i ng 39 below horizontal
x' ,max
from the southwest . In this upper hemisphe re repre sentation, i ts op-
posite i s the direct i on of the external shear force on the plane ,
as shown in t he inse t to f igure 6-38. The magnitudes o f TX r ,max
and a are obtained by reading the angle bet ween P 1 and Xl , which
f
x x
is here 24 . Then lax t / = 6.8 cos 24 = 6.2 and IT x I ,max I = 6.8 sin
0 0

0
24 = 2.8. The s igns on the ste r eonet can be inte r preted as follows.
Px , produces compress ion i f i t mak e s an angle less than 90 o with x" ',
which must be visualized as the outward normal to th e wedge. Sliding
of the contiguo u s blo ck wil l tend to occur down the face of the wedge
if TX :
, max is in th e upper hemisph e re.

CONCLUSION

Th e constructions u t ilized in this chapter are basic vector


G odma,,- Geo ' og ical _ng - 19
276 Applications of Stereographic Projection

Figure 6-38. Resolution of NORTH


stresses on the x' plane.

A
A

Upper Hemisphere

Direction of forces of
stresses on x' plane

operations. The illustrations are pertinent to stability of blocks,


and resolution of stresses; but the methodology can be followed for
other problems of rock masses involving vectors. Among these are
water flow through jointed rock systems, analysis of measured rock
movements, interpretation of geophysical measurements, and other
fields yet to be identified.
physical o els

Since structures in discontinuous rock masses usually provoke


opening or shearing of some discontinuities , almost every rea l rock
engineering problem is too complicated for closed form, mathematically
based calculations. As noted in the introduction to chapter 6 , rock
behavior can be assessed if the most l ikely modes o f failure can be
identified correctly. simple conceptua l models are useful for this
purpose. Once identified, the modes of failure can be weighed
experimentally in scaled physical models, or computationally, using
numerical models. Numerica l methods are the subject of Chapter 8.
In this chapter we wil l consider physical model methods, by means of
which the behavior of discontinuous rock masses may be explored and
extrapolated to prototype dimensions and conditions.

KINEMATIC MODELS

The word prototype refers to an i dealization o f the field problem


in which only those factors considered essential and relevant have
been retained. In the physical model study , the prototype will be
duplicated at a convenient scale with a minimum of distortion with
respect to the more important properties. I t is good practice to
experiment first with simpler, distorted models to determine the
essentials which must be duplicated. Such a preliminary kinematic
model study , for example , may observe the changing modes of behavior
218 Physical Models

.
pi
0 tR ~
~
0
- - ~
+

- CI\
- - -
~

'"
'"
'" !?
\!l

'" '"
; ~
~ -
::a
.,
Rig'" Toe _
Rlglil Toe

LeO Toe

0 .,"' .. ":'
~
:B
'" ~

- - '" '" '"


«]

'" " '"


~
'"
0
2 - - - - - - -
POWE R P L ANT Cf-t AMBER
!.CALI
I
.I

Figure 7-1. A geological log of an underground excavat ion prepared by Corps of


Engineers geologist James Zelti nger (Omaha District) in su ch a way that it transforms
into a geological model of t he discon ti nuities w hen ro lled into t he proper shape. T he
method of preparing such logs is described by Jack (1969 ). Court esy of t he Corps of
Engineers.

corres p onding t o d i f fe r e n t joint i n g s t yles as determined by geological


mappin g and study of geo l og ic al models (figure 7- 1 ).
A usef u l ma t e ri al f or simpl e t wo- d imensional ki n ematic models
o f exca vati on s i n d isco n t in uous r ocks is a mixture of flour, cooking
oil , a n d s and, s imila r to chi ldr en's "Play Do u gh." Such a material,
mixed to be l ow i t s p l a st i c l i mit , comb ines c u ttabi li t y and rigidity.
When smoo t h ed out into a shee t 1/ 4r! t o 1/ 2" th i ck, l i n es of discon-
Physical Models 279

tinuities may be pressed or cut . Each new cut c ompresses the mode l
by the width of the blade so tha t it is possible to produce a variety
of Jointing styles -- open or closed, planar or imbricated -- by
programming the cutting sequence . According to the oi l content, the
mixture may be made relatively plastic or brittle J and heterogeneous
structures can be modelled readily.
The main requirements for such a study concern cost and facility,
s i nce it is by repetitions and parameter variations that the full

(a) (b)

(c)

Figure 7-2. Procedure for conducting a base shear kinematic model study of a gravity loaded rock structure.
(a) Using a trowel , smooth out a sheet of model material
(b) Give the sheet a push to break the bond along its base.
(e) Cut the outline of the excavation (in this example, a rock cut with benches) and the system of
discontinuities.
(d) Push the model against the direction of gravity. The slope is failing by toppling, with the lower limit of
toppling defined by the discontinuity inclined towards the free surface .
280 Physical Models

spectrum of d e formatio nal mode s can b e studi e d. A mixtur e of o il,


sand , and fl o ur is ch e ap, universally avai l abl e , and imme diatel y
reusab l e. Howe ver, an y mi xture of i n gredi e nts which is p ressed in to
its f inal consistency ra t h e r t h a n harde n e d, e . g., b y c oo lin g o r
ceme n t ing, can also b e used e a si ly in s uch e xperi me n t s. Precut b l ocks
o f p las ti c (Trollop e , 1 966), s ugar cubes , wo o d , and cork , h a v e be e n
u s ed advantageously.
F igure 7-2 shows th e s e quenc e of p r e p a rin g and runni n g o ne
mode l e xperiment, i n t hi s cas e unde r si mu late d g ra v it y lo a d in g. Th e
e xperi me nt is conduc te d by push i ng t h e p r e p are d mo d el along a ro ugh-
e ned s u r f ace. Each e lemen t of the mo de l feels a she a r f o r c e along
i ts bas e which "foll ows" t h e d e forma t ion. The lowe r e d ge of t he mo d e l
repr es e n ts a l ine of c o nstan t e le vat ion, i. e. , stationary po i n t s on
t h e base surface are e f fec t i v e l y "mo v ing d ownwa r d " as t h e model i s
push ed upward. Alt e rn at ive ly , t h e bas e may b e move d whi le the mo del
is res t rained, as i n demon~trations c o nce i v e d by Dr. Eve r t Hoe k*
conduc ted on the surfa ce of an ove r h e ad pro j e ctor an d di s p l a ye d on
a p roj e cti on screen. Fi g ure 7-3 s h ows a l arg e si mpl e and i n e xpe ns i ve
"base f riction tt modelling machin e s imi la r to ma c hin e s at I mp e ria l
College, London , and a t Go l der and Brawn e r ' s o f f ices in Vancouver .
A con t in uou s sand pape r bel t is dr i v e n at cons tant ve l oc i t y by a
smal l mo tor, cre ating the gravity e f f e ct i n t he model wh i c h is
r e s tr a i n e d from following t h e sand pap e r by a fixed ba r ri er.
For e xperiments wit h excava tions lo ade d by b oundary press ures
where th e loading does not p ursue t h e d eforma t ion s , the mo d el ma y b e
loaded by applying su it abl e t hrusting p r ovisions in the sty le o f t he
p rot ot ype. It is als o possib l e to comb in e gr a vit y and non-fol l owing
di recte d l o ads, although t he comp lexity of t h e require d set - up may
defeat t h e basic premis e of t h e s upp o s e d l y s imp l e kin e mati c a nalySis.
Re l e ase of initia l stres ses cannot eas ily b e mode lle d , b u t t he re l a-
t ive ef f e cts and behavi or styl es f ollowi n g ex c avati o n in media of
differe n t rati o s of hori z ontal t o v erti ca l i ni ti a l s t r e s s may b e
cons i dered by pushing in a dire ction o ppos it e t o t h e v e c t or s um of
the vertical and horizontal init ia l n o rmal st r e sse s ( Goodman, 1973) .

* Innaugural Lecture , Impe r ial College o f Science and Te chno logy ,


Lon dOD , Feb. 1971 .
Physical Models 281

JOi n t or
Fa u lt Surfa ce Mode l ing
Mal er lo f

For ce Ind uced by fricti on


bel ween mov i ng su rface and (a) (b)
reocll on wil h fi xe d barr ier

Figure 7-3. (a) K inem atic model machine used by Go lder-Brawner Association to
stu dy strata movement in underground m in ing ventures. (b) Kinematic modelling
mach ine in th e rock mechanics laboratory at the University of California, Berkeley ,

Suppose t he be lt o f th e appa r a tus in figu r e 7-3 is d r iven at a


constant vel oci t y v, small enough t h at in er t ia forc es are n e gl i gibl e.
I n a mo del s ubjecte d to t rue grav it y , g , a fa ll ing block moves a
2
di st anc e s = tg t in t i me t. In t he b as e f r iction mode l in ti me tb
an unsupp o r ted b l o ck "falls" a dis tan c e s = vt b , The displaceme n t
at t ime t i n th e p r oto t y pe is , th ere f ore , cor rectly i n dicated by t he
base frict ion model at time t b g i ven b y

( 1)

Dr. J ohn Bray* f o un d t h i s s i milit u de r eq uire me nt a lso t o b e c o rrec t


f o r t h e case of a b lock sl i d i n g in a given d i rect i on o n a uni f orm
di sco n t i nuit y, If t he discon tin ui ty is inclined a t an angl e i > ¢,
t h e grav i t y loaded mo del a cce l e rate s downslop e, a cqui ring dis p lace-
2
ment set ) = !g (s in i - co s i t an ¢ )t , In the base f riction model ,
the blo c k and its s upport move d if fe ren tia l l y with th e relative
vel o ci ty vector ~, dire ct ed par all e l to t h e r es ultan t forc e ~ across
the di s con ti n uity ( f igu r e 7- 4) .

* Unpublis he d n o t es on simi l it ude in t h e base f ri c t io n mo de l, March


1973 , Impe r i a l Col le g e , Lon don.
282 PhysIcal Models

Figure 7-4. Mechanics of base f riction model of a block on an inclined plane .

..u = downsl ope velocity of bl ock


...V = sandpaper belt ve locity
V' = relative ve loci ty of blo ck
a nd sup port

The velocity vector triangle (figure 7- 4) gives

v u
sin(90 + <1» sin(i - <1»

or (2 )
v sin(i - <1»
u v ( sin i-cos i tan ¢)
cos <1>

I n time t b , the block subjected to base shear moves downslope a


di st a n ce s(t ) = v(sin i-cos i tan ¢)tb. Equating displacement i n
b
base shear and gravity models, we again find the result of (1). We
lack such a simple result for similarity in cases of overturning,
although qualitatively the model r e sults appear to be defensib l e.
The base shear method cannot duplicate the correct response when the
moving body acquires horizontal momentum since there is no mech a n ism
Physical Models 283

Constant
rotation

constan~
velocity/'1
~
I Gravity
ecce!. No
ho,.izontal
velocity
No
rotation

Gravity Model Base Shear Model

Figure 7·5. Dissimilarity of base friction and gravity loaded models where momentum
is not negligible; after Dr. John Bray.

for persistence of translation and rotation after "impact, rr or change


in friction of the surface of sliding ( figure 7-5).
Figures 7-6 through 7-9 present examples of simple kinematic
models in which deformational modes o f discontinuous rock masses
were examined. Figure 7-6 shows bas e shear models of rock s l opes
while figure 7-7 presents underground excavations loaded by the base
shear technique. The latter figur e i llustrates potential usefulness
of this modelling technique in comparing shapes for underground
excavations . Figures 7-8 and 7-9 illustrate kinematic models loaded
by pushing the structure i n the dire ction o f the vector of net load
on t he structure. The embankment dam models in figure 7-8 display
foundation cracking , opening of joints, rotation and flexure in the
foundation, and cracking inside the embankment. The arch dam case,
f i gure 9, is crudely simulated by pushing the whole arch directly
downstream , with a concentrated push at its center ; i t demonstrates
wedge sliding and f l exural failure modes. In the case o f an actual
dam, cracks on c e formed may f il l with water at some high percentage
of the reservoi r head, further damag i ng the rock mass. This is hard
to model kinematically .
In making us e o f the results of kinematic model exercises , it
must b e understood that t he behavior modes realized in the models
284 Physical Models

may be quite unlikely in the actual structure. The business of


assessing and weighing the various possible behavior modes requires
attention to dynamic simil i tude. Whereas, kinematic models are made
in an attempt to understand the problem better, the physical models
will be studied in an attempt to solve it.

(c)

(a)

(b)

Figure 7-6. Kinematic base shear models of rock cuts with two systems of discontinuities.
(a) A cut with horizontal benches, before pushing.
(b) The same cut as (a) after pushing. Sliding of a wedge occurs, causing overturning of
loose blocks under the benches. Note the caves between joints inside the loosened
rock mass. If the wedge were to stop moving, for some reason, these caves would
be preserved as evidence of the previous slide movement.
(c) Another wedge slide in development, causing caves to from between joint blocks.
Physical Models 285

PHYSICALLY SCALED MODELS

If dynamic sintlitude can be close l y approache d, a model can


reproduce deformat i on and failure of a discontinuous rock mass, With
regard to excavations and foundations in rock masses; we are interest -

(a) (b)

(
(c) (d)

Figure 7·7. Kinematic base shear model showing the effect of excavation shape on the
deformation mode .
(a) Flat roofed opening with hor izontal bedding before "turning on gravity" by pushing
the model.
(b) Classical symmetrical flexure pattern ; the upper lines of flexural cracks extend
through eight layers, with tolerable bending in the layers above.
(c) Further propagation of cracks and the beginning of a fall from the roof.
(d) An arch shaped roof of the same span; stable.
286 Phvsical Models

(a) (b)

(c) (d)

Figure 7·8. Kinematic models of earth dams o n bedded and jointed foundations.
(a) Horizo ntal beds in the foundati on close as the dam is pushed in the direction of
the load tra nsmitted t o t he fo u ndation by the dam and reservoir. T ension cracks
form in the upstream port ion of the dam fo u ndation a nd the differential defor·
mation of the foundati o n causes crac ki ng in the embankment.
(b) A similar case, except t hat vertical jo int s in the upper part of the foundation
accommodat e the down stream defo rmat ion with o ut new tension crac king. The
dam does not crac k, perhaps because the add itional joints permit shear deforma-
tio n in the foundation which red uces the shear stress at the base of the embank·
ment. Some rotation of blocks occurs in the upstream part of the foundation.
(c) A rock-f ill da m wi t h a shea r key o n a roack mass w ith beds dipping steeply
downstream .
(d) Case (c) after push ing the dam in t he d irecti o n of the load transmi tted to the
foundation by the dam and reservoir. F lexure of t he layers in t he foundation
has occurred, lifting the d am in the downst ream part of the toe trench and in
the downstream shell as the embank ment rotates a bout its lower m id dle region.
Physical Models 287

Figure 7-9. (a) A kinematic model of an arch dam in a canyon with vertical beds
striking into the left abutment and slightly downstream. Initial condition; the arch
is made of material stiffer than the model material. (b) First stage of deformation
as the arch is pushed downstream . Shear failure of the edge of one layer has taken
place on the left abutment and flexure of the downstream layers has initiated. On
the right side, a tension crack has severed the layer under the thrust of the arch.
(c) Continued pushing of the dam downstream has increased the flexure and
initiated flexural cracking on the left abutment, while sliding of a wedge defined
by the bedding and tension cracks occurs on the right abutment. (d) The dam
swings around an axis on the right side as the flexural cracks propagate on the left side.

(a) (b)

(c) (d)
288 Physical M odels

e d in s imil i tude with r espect to s tren gt h s, de formabi lities , l e ngths,


a nd t ime. Similari ty r e quireme n ts can b e e xp ress e d i n t erms o f two
i ndepe ndent d i me ns ion les s r a t ios (Fumag a l li, 1968) as f ol lows:

stre ngth o f proto type


(3 )
st rength of mo de l
and
length of prot o t ype
(4 )
l e ngth o f model

Th e scali ng r a t i o f o r a ny quantit y ha v in g t h e d i men s ions of f or c e


p e r uni t area is f;. The s cal i.ng r ati o for a n y qu an t ity having
di me n si ons o f )en gt h i s A. It f ol lows t h a t the s c ale r at io p f o r
q u a ntities wi t h d i me nsi o n s f orc e / vo l ume (dens it y) will b e

p ( 5)

t h e s c a l ing ratio ~ f o r f o rces wi ll b e

(6 )

a n d t he scal e of t ime

T ( 7)

where a is t h e s c a lin g ra ti o o f a cce l erations ; a wi l l be un i t y un less


a c ent ri fuge is use d. Dime nsion less q uant i ti e s l ike Poisson 's r ati o,
an gl e of f ric tion , an d st rai n s h o ul d b e t h e same i n model an d
pro to t yp e .
It is no t possib le with a g i ven mat e r i al t o achieve t h e req u i r ed
s ca l i ng r atio for a l l quanti tie s . F or e x amp l e, th e leng th scale is
f i xe d on ce th e mo del s ize is se lecte d. It may b e ne c e ssary to
e s tabl ish this s c al e main ly to us e e xisting eq u ipme nt . On ce f i xed ,
the l ength s cal e g ove rns o ther le ngths s u ch as t h e scale o f she a r
d isp laceme nt at f a ilure , and th is may or may n o t b e obt ainabl e wi t h
mat e r i a l which has b e en ch o s e n t o meet o t h er re q u i r e me nt s . If t wo
qua nti ties o f the protot y p e whi c h have th e s a me di mensio n s a re i n
Physical Models 289

f ac t not equally sc a le d, the mo de l i s said to b e d i s t orte~ . A certain


amoun t of d istort i o n is i n evit a b le . As another example, both t he
u nconfined compress i ve stren g th an d t he modulus of e lasticity have
dimensions of force pe r un i t a re a. Th e ratio of modu lus to stre ngth
for many rock s is of the order of 500 ( Deere et al 1967) ; in t h e case
of the flour-oil-sand mixture me n ti on e d e a rli er, t hi s rati o is o f th e
order of 4 , so eit h e r the d eforma t i o n s or t he s t re ng t h will be dis -
t o rted. The cho ice of ma t er ial s mus t repres ent a comp rom i se i n
wh i ch all factors b e a r ing on t h e r e s u lts are considered and weighted
in impo r tan c e a nd t he n ecessar y dis tortions rele gated to th e l east
i mp ortant quantities. For this r eason t here c an no t b e one modelling
material fo r all pur p oses .
In ma n y cases, rock deformations a re of f irst inte rest , for
e x ample i n the p robl e m of fin ding t he stresse s i n a dam resulting
from displacement and rot ation of i ts rock abutments. If al l rock
deformation is i n the elastic region , primary att e ntion in selecting
a mater i al wil l be a d dresse d to s ca lin g i ts modulus o f e l a st icity
an d Poisson's r a t i o. Sc alin g t h e de fo rmat io n o f th e d isco ntinuiti e s
necess i tate s choosing t he widt h o f a p e rtur e to give t h e correct
va lue of ma xi mum c los ure V and the co rrect shear stiffness (see
me
Ch apt er 5 ). Howe ve r, as disc us se d by Barton ( 1 972), the peak shear
d ispl a cement i n di rec t she ar tes ts i n creases wit h samp l e size, while
the pe ak strength is i n depe ndent o f sample size . Therefore , t he
shear stiffness disp lays a stron g scale e f f ect , and the mode l wil l
g i ve d i storted scaling f or al l but on e sc a l e ratio .
Whe re pl a st i c behav i or of t h e dis con t inuities or the rock is
pre di ctable, as i n st u dy o f r o ck failure pr o ces ses , t he material
will b e sel ect e d to scale ~ a nd c an d t h e c omp ress i v e and tensi le
s trength s, wh i le f o r t he disc on tinuit ies t he peak and re s idua l
f ri cti on an gl es and t he d i latancy wi ll become t h e pr in cipal object s.
Stimp son ( 19 70) reviewe d model lin g mate ria l s used in rock
mechani cs . Gran u lar mater ia ls s uch as cemen t and plaster are
dilatant while p lastics tend to b e n on-d i latant. Tables give t he
unco n f i n ed c ompress ive a nd tens ile st rength , e lastic properties, and
u l t i mate strain f or many mate ri als . Ma t erial s us e d to represent
ro ck in mo del te sts in t he non-elas tic domai n in cl ude: cement ,
290 Phvsical Models

sand and water; san d, wax, and mica; sand and clay; and plaste r
alone or mixed with sand, clay, mica, barite, lead oxide, d iatomit e ,
s awdust, or l i me. Sand and wa x mixtures tend to be plastic wh ile
p las ter or cement s and mixt u res are brittl e . Strength, def ormabil ity ,
a n d u n it weight can be va r ied ove r wi de ranges by controlling c uring
an d mi xi ng time, and mixing additional materials. For example, sand
plaster mixes are weakened by curing at 90 degrees C, or by addi n g
crushed mica; they are made more brittle with chalk or clay a dm i xtur e;
a nd adding powdered lead or barite raises their specific gravity.
Joints and other discontinuities have been simulated in man y
ways -- as rough extension fractures caused by b reaking the soli d
modelling material, as untreated or varnished saw cuts in the s olid ,
an d as thin partings of wax, grease, talc, limestone dust, wax paper ,
oiled tracing paper, lens tissue, g r aphite, and clay. It is e asy to
duplicate t he low friction and re a dy parting of important se ams a nd
faults, e.g., by grease partings, b u t it is harder to raise the
0
friction angle above 4 0 and to control dil a tancy. Krsma n o vic and
Associates developed the technique of imbricati ng joint blocks, i.e.,
offsetting blocks slightly to produce block interlocking. The
resulting joints show high strength and dilatancy at low normal
pressures but they are highly directional i n their shear be havi or.
Barton (1972) developed an indirect tension splitting device (te rmed
a "guillotine") to introduce rough extension fractures in cast pl a tes
o f model mate ri al. In contrast to cut joints, the oldest set of
s uch split discontinuities is the only one which is planar, sin c e
all younger discontinuities are offset wherever they cross t h e
earlier sets.
A simple, plane stress model study of a hypothetical un de rg round
power station has been used for several years as a class exerci se
in the rock mechanics course at Berkeley. The model, at 1:150 len gth
scale, is prepared in about one half day by placing hot mixtures of
sand, crushed mica, an d paraffin (2-4%) according to the desired
lithologic section. The model apparatus consists of a wooden b o x
o f ins i de dimensions 30 inches wi de by 25 inches high b y 4 i n ch es
t h ick (f i gure 7-10). To construct the model, the cross section i s
drawn to scale and attached to the plywood front piece (figure 7-11a).
i+i
tl i+i
rl
i-o-i
ll
l'*i
s
o
e
3
tl
!qaFl i+i i+i ;qF'
o
I
o
l@i
:r
@i i+i i*i E
o i+i @i i€.i ic€Fi R
l@ 2 o
6'
8r
i@
l@'l
tl
i+i
4, s
m
F
I
o
i+
t@
i+
,<: _ i@i
ll
i.@-i

i+
i+i \ i*i
i+i i+i
BACK VIEW DURING CONSTRUCTION BACK VIEW DURING TESTING
tlrlrtl
o lFf.

NOTES'

l. Ply.wood fronl during consf ruction of model ;


reploced by plexigloss for testing.

2. Bottom 4 plywood cross pieces ore reploced, ofter


construcfion of model, by o single plywood piece con-
toining o window in the shope of the eventuolercowtion.

5. Wood blocks lo simulole moferiol to be excovoled:


embedded in the modelond withdrown throuqh window'
in bock, under lood. Alternotirply wood bloc-ks moy be
omiffed ond model rnoy ocluolly be excovoted.

4. Hordwood posts

5 Plywood bose plole


FRONT VIEW DURING TESTING END VIEW

Figure 7-1O. An apparatus for two dimensional scale model studies, u


292 Physical Models

Figure 7-11. Scale model study using the apparatus


of figure 7-1O.
(a) Model during construction, viewed from the back.
(b) Before testing, viewed from the front.
(c) After testing to failure. (c)

The back of the model is installed as the model is built up; then
the plywood front is removed, a grid of reference points is sprayed
over the surface, and the clear plastic front is attached. To
simulate the excavation sequence, hardwood blocks having the shape
of each excavation stage are buried in the model (figure ?-Itb) and
later withdrawn under toad, through a hole in the plywood back.
Simple studies such as this are useful for many facets of
engineering for underground and surface excavatj-ons, for example:
identifying the critical points of an excavation; studying variations
of shape and locatlon; developing rational bases for designing
supports; and putting in context the deformation readings of individ-
ual instruments . Ivluch more elaborate studies are possible, of course,
and in fact may even serve as the basis for quantitative design
decisi-ons. Figure 7-l-2 is a typical result f rom many coal mining
Physical Models 293

Figure 7-12. Flexural failure of a laminated mine roof; a model studied by Dr. Everling
at the Bergbau-Forschung, Essen, West Germany. Courtesy of Dr. Everling.

studies in the Ruhr district of Germany.


Two dimensional models of long structures such as tunnels, shafts,
road cuts, and long-wa1l mine faces will usually be more closely
correct in a plane strain configuration than in plane stress. There
is no fundamental difference between them in elastic problems, but
in models carried into the rupture region there may be great differ-
ences between plane strain and plane stress. f'ailure will tend to
occur in and along the discontinuities in either case, but new
fractures in the rock will propagate in the plane of least principal
stress. In the plane stress mode1, if there is no tension stress,
the material will crack tongitudinally parallel to the model p1ane.
To prevent such cracks, plane strain conditions can be approached by
providing stiff, constrained front and back plates. Unfortunately,
294 Physical Models

this introduces significant friction as weII as limiting access and


observability.
Three dimensional models, more difficult to build and to
instrument, are justifiable for many problems which are truly three
dimensional since no other engineering method of analysis can yet
duplicate complex kinematic, geoloBic, and structural requirements
in three dimensions. Examples of three dimensional problems are
caving and subsidence above mines, roof conditions in underground
intersections, condj.tions at the face of a tunnel, an arch dam and
its foundation and abutments, and an open pit mine or spillway
exeavation. Further, almost any anisotropic rock mass necessitates
a three dimensional analysis as only rarely will the planes of
elastic and shape symmetry coincide. The following illustrates a
careful three dimensional model study of a problem in discontinuous
rock.
Example-Grancarano Dam, Yugoslavia
A model of this I2O meter high arch dam and its bedded and
jointed foundation was made by Prof. Krsmanovic and co-workers Langof
and Tufo of the faculty of Civil Engineering, Sarajevo. The lime-
stone beds, O.2 to one meter thick, separated frequently by clay
partings, dip gently upstream and into the right bank (figure 7-I3).
The rock is also extensively jointed with smooth and rough unfilled
discontinuities and has calcite and clay-fiIled minor faults.
Statistical analysis 1ed to a simplified picture with two joint sets
and additional important individuals in other orientations. Joint
and. rock properties were measured in the foundation excavation using
f lat jack and field shear tests (Sr2 area) ancl geophysj-cal measure-
ments. Eleven homogeneous zones were defined with the modulus of
deformability in the range 7,600 to 21,500 Mw/m2. Samples of beds,
joints and minor faults of the rock were also tested in the labora-
tory, resulting in classification of the discontinuities into several
behavioral types with friction angles from 15 to 25 degrees for clay-
filted seams and considerably higher for rough, unfilled joints. fn
the latter case, g varj-ed continuously with normal stress.* Properties
1
* O is defined at any normal stress (o) as tan-t(r-/o). A "c O
material" by this approach becomes a material invwhich 0 varies
with 6.
295
Physical Models

Figure 7-13. Bedded limestone in the left abutment of Grancarevo Dam, Yugoslavia;
courtesy of Dr. M. Popovic, Institute for Geotechnics and Foundation Engin., Faculty
of Civil Engin., Sarajevo Yugoslavia.

of the intact rock were established as follows: E = 80,000 to


,.ug,, = 1oo to 1bo Mt'I/mz, and specif ic gravity = 2,5.
loo,ooo MN/mz,
The scale ratios selected were:
-

296 Physical Models

I = 80 (the ratio of lengths)


p = 1 (the ratio of unit weights)
gr_vr.ng E = 80 (the ratio of stresses ).
The material selected for the dam actually had a lower density than
required; additional weight was added externally at distributed
points. The material selected for the rock blocks a c1ay, plaster
mixture had values of E = 11OO UN/mz, e,,
'u = L.4 Mtt/mz , and specif ic
gravity 2.5. The latter high value was achieved by admixing lead
grains
Thirty three of the more important individual discontinuities
in varying attitudes were represented with friction angles assigned
either 25 or 40 degrees. The recurrent joints and bedding were
approxi-mated by an imbricated pattern of b locks with an average of
2mm of offset per 9cm of length. Choosing one size of blocks fixes
the ioint spacing and would determine the deformability identically
throughout if the joints were identical. However, by varying the
initial normal compression and fit during model building, the initial
normal deformability could be adjusted within the required range.
111,OOO cubes 4cm on edge were built into the model. Clay layers
were duplicated by grease partings, givi-ng 0 of about L4 degrees.
Continuous variation in bedding plane attitudes was not realized, but
the volume of the model was subdivided into seven zones of constant
dip; partj-al cubes were placed to complete the space at zone boundaries.
For construction, poj-nts were located in space using plumb lines
suspended from a plan placed above the model (figure 7-L4). The
dam was poured in place and loaded by hydraulic cylinders on the
upstream face. Body forces due to seepage in the abutment were not
duplicated in the model but to simulate uplift in the foundation,
a rubber pillow was inflated inside the rock mass under the dam
upstream of the grout curtain. Instrumentation consisted of pressure
sensing blocks emplaced. in the rock mass and strain gages and dis-
placement points on the rock and dam surfaee.
The investigators learned that the rock mass took less time to
adjust to load or unload increments than did the dam itself, pre-
sumably due to the dam's higher stresses. First the reservoj-r load
was applied and removed quickly; then, b€fore proceeding with the
Physical Models 297

(a) (b)

Figure 7-14. (a) View of the model during construction; note the use of plumb lines
to locate points in plan; courtesy of Dr. M. Popovic, Sarajevo. (b) A gener al view of
the model; courtesy Dr. M. Popovic, Sarajevo.

Figure 7-15. Lines of equal


deformation on the rock
surface under 3.5 times the
reservoir load. Displacements
'Y =3_50 in mm. at the prototype scale.

next level of loading, the load was s us t ain ed for 5 to 10 d a y s.


Figure 7-15 shows lines of equal dis p lacemen t of po in ts on t he r ock
surface at a loading corresponding to three an d a h alf t ime s that of
t he full reservoir (y = 3.5). The found a tion def o r mations i n the
lower left abutment were the object of a re inf orce ment s ch eme us i ng
prestressed cables as shown in figure 7-1 6. That part o f the
298 Physical Models

Figure 7-16. Reinforcement scheme to stiffen and stren gthe n t he fo undat ion
in the left abut ment. It was designed on the basis of the results of the mode l st udy.

foundation conta ins unfavorably or ien ted strata a nd cl ay seams .


The res u lts of this comprehensive model st u dy evaluate d f or t h e
designers what the reserve of strength was likely t o b e an d s h owed
where the weaker points lay in the dam and in the foun dation . The
studies also det ermi ned the size of the volume o f ro c k par ti cip ati ng
in the s tr u ctur a l behavior of the d am un der full reservoir load.
The model i ndicat ed where impro veme n t coul d be ma d e in de s ign and
Physical Models 299

where the rock might be strengthened. And it provide d a framework


for evaluating t he response of instruments in the actual structure .
Only a physical model , presen tl y, can provi d e this degree of knowledge
about a. struc ture on discontinuous rocks. The principal d i fficulty
in such a study is that aft e r tests to rupture, it cannot b e repeated
readily with diffe r en t paramet e rs . Numerica l models which have
this capabilit y, may eventually be able to duplicat e the three dimen-
sional rock details as faithfully as t he physical model .
8
the finite element method

INTRODUCTION

Finite element analysis is a digital computer method for stress


analysis and other tensor field problems of large size. It is
especially powerful for non-linear rock problems introduced by
heterogeneities and discontinuities for which closed form solution
methods are difficult and special.
Not all problems require a finite element solution. Problems of
elastic continua in two dimensions, and some three dimensional cases,
can be solved for many boundary conditions by direct application or
superposition of published results of the theory of elasticity.
Limiting equilibrium of blocks with simple geometry in slopes for
which intermediate stress states are of little interest, are more
readily solved by methods explored in chapter 6. Finite element
analysis is immediately useful, on the other hand, in situations
where displacements need to be known, particularly in heterogeneous
or discontinuous rock masses. This need arises when interpreting
deformation readings in-situ, when scaling up model experiments or
extrapolating prototype studies, and when studying the action of
r ock foun da tion movements on the behaviour of structures. A number
of references illustrating various applications of finite element
analysis are given in the bibliography.
The Finite Element Method 3G1

THE METHOD

The f i nit e e leme nt method is an app l icati on o f t h e di r ect


stiffn e ss method of structural analysis. In th i s approach, di s -
p l aceme nts of r ep resent at ive poi n ts wi t hin the structure , t erme d
nodal points, are the variable s of a s e t of simul t aneous e quat ions.
The coe fficien t matr i x *, describing the geome tric a l and physical
propertie s of the structure , is termed the s t ru c tura l stiffness
matrix .
"S t iffnes s ' me a s ur e s the amo u nt o f for ce n e c e ssary to produc e
!

equilibrium in a bo dy unde rgoing differ e n t ial d isplacement s . It


i s a system proper t y , since it r e flects bo t h the s h ap e and si z e
of the body and i ts physical properties. Consider f or examp le, the
stiffn e ss (k) o f a coi l sprin g, defin e d by W = k( ~ u) , whe r e W is the
orce stretchin g the spr i n g a n d 6 U is the re s ult i ng s t retch; k dep ends
not onl y on the kind of steel us ed, but on t he cross s e ct i onal area o f
t he wire, and on its coil c onfi g ura t ion (figure 8-la). S i mi l a rl y , as
shown i n figur e 8-lb , t he compre ssiona l s t i f fness of a cy l indrical
specimen of e las t i c rock depends not onl y on i t s modu l us of e la s ticity
(E) but on its c ross sectiona l area (a) and length (L).

aE
k =- (1)
1 L

When a body has multip l e f r ee doms t o de f orm i n various modes , one


can di v ide t he whol e body into sub-element s and develop a st ruc t ural
stiffn e s s matrix f rom the individual sti ffnes s components of e a ch e l e-
ment . In t h e direct stiffness method, after the st ruc t ura l stiffne ss
matrix is ass e mbled, speci f ic fo r c es and constrain t s are i n t roduce d to
obtain a specific tl solut ion", that i s , t he s e t of di s plac e ment vect ors
at each nodal point. The n the forces or s tresses in each compon e nt
are det e rmined by mu l t ip ly ing t h e k nown displace me n t s by t h e known
stiffne ss t erms f or each eleme n t.

*The no t ations and operations of matrix algebra s implify t he ma t erial


of this Chapter. Matrix a l gebra is d i scussed by Wy l i e ( 1960).
302 The Finite Element Method

~~~~~~~~~~~~~/~I

'" . ,
. <
.
,.

)
--I,
rock:
~ ~u

modulus of elasticity = E
'. -, . cross sectional area = a
- length = L
.
"

"
", ~. • .,' \ . I.

~u
, .
--L __ _

W = k ~u

(aJ Spring stiffness (bl Rock in co mpression

Figure 8-1. Concept of "stiffness".

A simple one dimensional example will illustrate the essential


steps followed in the finite element scheme.

Example 1

The structure in figure 8-2 a, consisting of a column of two


different rock cylinders is loaded by a force P. Find the displace-
ments and stresses throughout the column when the bottom is moved
upward by an amount y and held in that place.
Step 1. - subdivide the structure into idealized elements
with a finite number of nodal points as shown in
figure 8-2b.
Step 2. - form the element stiffness matrix for each element.
We rewrite Llu as VI - V : Then the external forces
J
at Nodes I and J of element A required by the
relative displacement VI - V J are
The Finite Element Method 303

I
r
1

Fext I

Rock cross sectional A


- 1 - ,--...... node I
area = 0 1
Fint I
J
2
I I
Fint J
Rock ~
_ 2 -
B ~ node J
F
extJ

BoUn dary } u- [ cl An eleme nt in equil ibrium


cond ition 'S" y under external compression.
(a) Prot ot yp e
Internal forces positive at 1.
(bl Idealised st ructure and neg ative at J
2 elements
3 nodal pOints

Figure 8-2. Simp le illustrative problem-example 1.

The s ign conven t ion use d re c ognize s f orces a n d dis pl aceme n ts


pos itive whe n po i nted u pward. The forces Fr
and F J ar e ext e rn a l t o
t he e lemen t. Their e xis t enc e implies a n equa l an d opposit e r e a ct ion
inside the e l eme nt . A downward a ppl i e d ext e rna l force P prod u ce s a
compress i on in e ach e l eme nt , wh i ch nece ss it at es i n te rnal fo rce s -F 1 ,
and -F J t endin g to stretc h the body back to its orig i na l lengt h if P
is removed, as shown in fi g ur e 8- 2c.
The e lemen t s t iffne s s matrix for e lemen t A is obtained by re-
writing equation (2) 1n . matr~x
. .
notat ~ on.
*

(3a)

* {A} means A is a column matrix whi l e <A> me ans A is a r ow matrix.


304 The Finite Element Method

Similarly, for element B:

( 3b )

We may term the column matrices on the left hand sid e "nodal po i n t
external force vectors" {Fe} while t h e righthand column matr i ces ma y
be termed !fnodal point d isplacement vectors!f {u}.
St ep 3. - Assemble the st r uctu r al stiffnes s matrix. The e x terna l
forces for all nodal points of the structure can now be e xpr e sse d as
the products of the nodal point displacements (un k nowns) and app ro-
priate stiffness terms. The stiffness term kpq gives the c ont ribu tion
to the force at nodal point p cause d by the displacement at no d a l p oint
q; it is the sum of all element stiffness terms linki n g p and q an d
therefore is zero if p and q have no elements i n common. For e xamp le
(figure 8- 2b), the increment of e x tern al force F2 at n oda l po i n t 2
produced by a displacement u 2 at noda l poi n t 2 fo r the stru ct u re
consi dered is k 22 u 2 where k2 2 = k~J + k~I = k A + k B·*
Considering each nodal point in tu rn we can wr ite

-k a
(4)

or since the in te rnal forces a r e t he negatives of th e ex te rnal f orc e s


(figur e 8-2c)

( 4a )

where: {F.} is the nodal po i nt in ternal force v e ctor for t he who le


1

* The double subscript IJ i n dicates t he r ow and column re spec tiv ely


of the term in the appropriate element stiffness matrix.
The Finite Element Method 305

structure. In a two-dimensional prob lem {F i } h a s rank 2n x 1, wher e


n is the number of nodal pOints ; {u} is th e n o d a l p oint dis pl acement
vector for the structure, with rank 2n x 1 i n a t wo dimensiona l problem ;
and (K) is the structural matrix, (defin e d to be pos i tive) , wit h rank
2n x 2n in a two-dimensional problem.
Step 4. - Introduce the external ap pl ie d forces and write the
equations of equilibrium. The ve c t or of e x t e rna l f o rces appl i ed t o
each nodal point may in clude contribu ti o n s from e x t erna l pressures,
water, accelerations, temperature eff e ct s, or othe r sourc e s. Initial
stresses may be included as external fo r c e s, as d i s cussed later . In
example 1, the only applied force is a downw ar d l o ad (-P ) at noda l
point 1. The net external forces (appli e d mi nus i nitial ) i nduce
deformations until the resulting intern a l f orces come into equilibrium
with them. This is equivalent to spec if ying, f o r a static probl e m,
that the sum of external a n d internal force s e q ua l s zero .

{F } + {F.} (5)
e 1
°
In the example being discussed, {F } = <-P, 0 , X> T wher e X i s the
e
reaction to the displaced and fixed b o u ndary c o ndit i on at nodal
point 3. Introducing this value for {F } and a l so substituting for
e
{F.} from (4a) yields the simultane o u s e q u ations of the structure
1

-P

( 6)
°
X

Step 5. - Introduce displaceme n t bo u ndar y c onditions. When any


nodal point displacement is specifie d , t h e l is t of unknowns is
. shortened. Thus we can set aside (part i t i on ) the row o f t he structural
stiffness matrix corresponding to the r ow of t he known displaceme nt .
Also since the stiffness terms are known , we c an t rans f er to t he known
external applied l oad ve c tor (on the left s i d e ), t he p roduc t s of known
stiffnesses and known displacements. This de l et es , fr om the structura l
stiffness matrix, the column having the same n umber as t h e row number
306 The Finite Element Me thod

of the known displacement.


Whenever a displacement is specified, an e x te rn al force at that
nodal pOint is implied. It is not necessary to so lve for th is force,
to solve the system, since the row involved is removed from t h e
simultaneous equ a tions. The unknown reaction can be found lat e r, if
desired, from the known nodal point d i splacements of the eleme nts
containing the s p ecified nodal poin t .
Return ing t o t he example, remove the t h ird e qu ition

x (7a)

Then si n c e the disp lacement of noda l poi n t 3 was specified e qual to


y as a boundar y cond it io n , move the products of y to the lef t - hand
side of what rema i ns of t h e system of e qua t i ons ( 6) givi n g :

(7b)

Step 6. - Solve for the unknown d i splacemen ts . A solu tion to


(7b) is possib l e if the determin a n t of the coeff i cient mat ri x is not
zero. Th e sol u tion is

and

P
u2 Y - k
b

All nodal poi n t s d i splacements ar e now known.


The Finite Element Method 307

p p
y - "k-
a kb
p
{u} y - (8 )
kb
y

Step 7. - Find the stresses in each element and the external


reactions at the supports. Knowing the individual element stiffness
matrices, multiplication by the nodal point displacements, (3a) and
(3b), yields the external forces. The stresses are related to the
forces in this simple case by ° = F/a so we can modify the element
stiffness matrices to yield stresses directly; taking tension stress
positive *

-p

where Fr and F are external forces in element A.


J
Similarly, in element B,

Fr -F -p
J
°B a2 a2 a2

This ma y seem a long process for such an obvious result, but it


becomes attractive with only a few elements since the matrix operations
are efficient on a digital computer.

FORMU LATION OF ELEMENT STIFFNESS MATRICES AND EXTERNAL LOADS

The steps outlined above are the same when a more general two
d imensional structure is solved; only the element stiffness matrix

*For finite element work it is a good idea to suffer the temporary


inconvenience associated with a tension positive sign criterion.
Even though rock is usually in compression, commercial programs,
us u ally based on computer programs written by structural engineers,
take tension positive. Errors, particularly in shear stress signs,
are likely to creep in from partial mixing of sign conventions.
Good man-Geolog ical Eng.-21
308 The Finite Element Method

requires further elaboration. Two kinds of elements are used in the


computer program presented in Appendix 1 orthotropic constant strain
triangular elements to represent the rock material, and linear linkage
elements to represent individual discontinuities. Many other types of
elements are used in finite element programs -- e.g. bars, shells,
quadrilaterals, axisymmetric solids, tetrahedra. But we will discuss
only these two to simplify the presentation and the program.
A general formula for the stiffness matrix of a finite element
is as follows (Zienkiewicz, 1971): ,
K (9)

where L is the matrix relating strains to nodal point displacements,


o
and C is the matrix relating stresses to strains. We will use this
formula to develop the stiffness of a constant strain triangle.

THE CONSTANT STRAIN TRIANGLE

The constant strain triangle, introduced by Clough (1960),


develops a constant state of strain throughout the triangular area
between three neighbouring nodal points as an approximation to the
varying field of strain actually occurring. In practice, where
strains vary over short distances, the triangles must be made small.
Figure 8-3 shows a triangular element in a "mesh" of contiguous
triangles representing a continuous part of a rock mass.

Strain-Displacement Relationship (Lo)

As a result of a deformation of the triangle (figure 8-3) an


interior point at co-ordinates x, y is displaced by amounts u x ' u y
given by:

u (x,y)
x

and

u (x,y)
y
The Finite Element Method 309

this can be written

a
o

lOx 0 y 0
(10)

or {u(x,y)} = (~) {a} (lOa)

This variation of displacement with position preserves the connection

Mesh before st raining Mesh after straining

x
I
j Figure 8-3. Constant strain triangles.
310 The Finite Element Method

between adjacent triangles no matter how large the strain, (figure


8-3). Since we are particularly interested in displacement at the
nodal points, apply (10) to each of these points in turn. The
result is simplified by taking a local origin at nodal point I,
an allowable step since the stiffness matrix of an element must be
independent of its position in the structure. This yields:

o o o o o
u y,I 1 o o o o
u x,J o o o
(11)
u y,J 1 o
u x,K o
u
y,K 1 o

(lla)

We now can obtain an expression for {a} in terms of the nodal


point displacements.

(12)

Performing the required matrix inversion gives:

XKYJ 0 0 0 0 0

0 xJYK - xKYJ 0 0 0 0

1 YJ - YK 0 YK 0 -Y J 0
(q, )-1 (12a)
0 xJY K - xKY J 0 YJ - YK 0 YK 0 -Y J

xK - xJ 0 -x K 0 x 0
J
0 xK - xJ 0 -x K 0 xJ

which can be confirmed by multiplying by q, o·


The Finite Element Method 311

The determinant xJYK - xKYJ equals 2a, where a is the area of the
element. The strain components in the triangle are:

au x
E
X ax
au
E --.X
Y ay

and
au au
x + --.X
ay ax

this may be written:

ux(X'Y»)
(13)
( u Y (x,y)

Substituting for {u(x,y)} from (lOa) and performing the differenti-


ations yields:

(1JJ){cd (14)

where

~)
o 1 o o
( '¥ ) o o o o (14a)
o o 1 1

Finally substituting for {a} from (12) in equation (14) gives:

{E} = ('¥)(¢ -l){u}


o
= (L 0 ){u} (15)

Since (¢ -1) was determined (equation 12a) the strain-displacement


o
matrix (Lo) of equation (9) is now determined.
312 The Finite Element Method

YK 0 -YJ 0
1
0 -x K 0

-x K YK XJ

(16)

(16a)

where (L l ) is the matrix on the right hand side of (16).

Stress-Strain Relationship for the Rock

The stresses inside the triangle can be expressed in terms of


the strains and the deformability properties of the material. Assume
the rock is bedded or banded with the direction of the normal, (n)
in the plane of the mesh, (figure 8-4). The trace of the layers in
the element is direction s, and the direction perpendicular to the
mesh is called t. The counterclockwise angle from x to s, in the
triangular element, is a. First we will develop the stress-strain
relationships in directions s, n in which the properties are most

(a) (b)

y
n
trace of bedd ing

s. n. local coordinate directions


x,y, global (mesh) coordinate directions

Figure 8-4. (a) Definition of local and global coordinate directions; (ex) positive.
(b) Relationship of local coordinates and elastic symmetry directions-transversely isotropic rock.
The Finite Element Method 313

naturally expressed; then we will rotate them to the mesh co-ordinates


x, y.
The relationship between increments of strain and stress is
presumed to be elastic. Denote by the term v .. the Poisson's ratio
lJ
that tells how much strain occurs in direction j due to a stress in
direction i. Then the strain increments are related to the stress
increments by:

-v v
sn 1 tn
E IJ.° s + IJ.o -E; IJ.° t (17)
n
~ En n

and v
-v
IJ.° n +~
st nt
E IJ.° s - E IJ.° t
t ~ En t

The stress-strain relationship is symmetric. If the material is


bedded parallel to s (figure 8-4) it is isotropic in the st plane.
The following relationships then hold:

v v
st ts

E
s v
vsn En ns
(18)
En
v v
nt Es sn

For plane strain, E 0, and (17) and (18) yield:


t
314 The Finite Element Method

~at = v s t~a s + v sn ~a n (19)

and the strain-stress relationship simplifies to

{s} = (D){~a} (20)

and:

2
-v
(1 - v st sn
) - E - (1 + V )
s st

2
-v sn v
1 sn
(D) (1 + v ) (20a)
~ st En -
~
0 0

The stress-strain relationship is {a} = (C){s}


where (C) = (D)-I. Performing the matrix inversion yields:

- En v 2sn En v sn (1 + v t)
s
1
(C) E (1 (20b)
m n

o o

and

2
m (1 + v s t) [1 - v s t - (2E n IE s )v sn ]

Anisotropic stress-strain properties may originate from inherently


aniso~ropic rock; or they may be introduced to describe the overall
The Finite Element Method 315

behaviour of rhythmically bedded or banded rock. In this case the


terms v st ' v sn ' E s ' En' and Gsn can be calculated from the deform-
ability properties of the rock and the discontinuities (Goodman and
Dunc a n , 1971).
For the rock mass depicted in figure 8-4b, with joint spacing
h, and isotropic rock:

1
En 1 +
E
1
k h
"
n

E E
s

(21)
1
G
sn 2(1 + v) 1
E + k h
s

v
sn

and v
st
= v

where k n and k s are the normal and shear stiffnesses of the joints
(discussed in chapter 5) and E and v are the elastic modulus and
Poi s s on's ratio of the rock.
The stress-strain relationship in the local co-ordinate system
sn can now be wr i tten:

{lw} (C) {E} (2 2)


s,n s,n

We c a n generalize this relationship to the global co-ordinates of the


me s h , x, y, by expressing the strain and stress increments in the
system s, n in terms of the strains and stress increments in the
system x , y a s follows:

s,n
316 The Finite Element Method

and {flO}
s,n

substituting these expressions in (22) and using

gives {Llo} (22a)


x,y

where (T E ) is defined by (Jaeger & Cook, 1969):

2 a. 2 1
sin
cos a. 2" sin 2a.
2 2 a. 1 (23)
(T ) sin a. cos -2 sin 2a.
E

-sin 2a. sin 2a. cos 2a.

For an isotropic body Es = En


t
=
= E, all Poisson's ratios = v,
E
E ~0
and Gsn = 2(1 + v ) ' The stress-strain relationship for pla~~ strain
simplifies to:

1
v 0
1 - v
E{l - v} v
(C)
(1 + v)(l - 2v) - v
1 0 (20c)
1
0 0
2(1

Element Stiffness Matrix


We may now substitute the strain-displacement relationship (Lo)
(16) into the stress-strain relationship (22a) and introduce both
into (9). Since all the terms of (Lo) and (C) are constants, the
integration yields only the area of the triangle, which as noted is:
1
a = 2 (xjYk - xkYj)' Then

(K) (24a)
The Finite Element Method 317

where Ll is the quantity defined by (16a), (defined i n glo b al


co-ordinates x, y). For an elastically isotropic rock

(K) (24b)

in which (C) is given by (20c).


The computer program performs the matrix operations directly rather
than substituting in an explicit formula obtainable by e x pandi n g
(2 4a). (L l )(3 x 6) is defined by (16) and (16a); (Ts) (3 x 3 ) is
defined by (23); and (C)(3 x 3) is defined by (20b). Si n c e (K )
relates 6 forces to 6 displacements, it is 6 x 6 in size.

INITI AL ST RESSES IN THE ROC K

In many practical problems with rock the initial stresses are


of comparable order of magnitude to the stress changes ant ic ipated
and furnish important contributions to the vector of applied loads.
Equ i librium under an initial stress state implies initial in ter nal
forces throughout each element resisting the p re ssure of a d jacent
elements. When new free boundaries are ex c av a t ed, the equ ilibrium
is disturbed and the previous set of internal for ce s deforms the
rock producing stress changes sufficient to create a new equilibrium.
The total stress at any time, (0), is defined b y:

{a} (25 )

where {o } is the vector of initial stresses referred to x, y.


o

o
x,o
o
Y,o
T
xY,o

The external force vector Fo in equilibrium with the initial stress


is:
318 The Finite Element Method

(26)

Substituting (Lo) from (16a) in equation (26), and observing that


the integrand is constant, the external forces equilibrating the
initial stresses are:

(27)

The net load vector on the structure due to applied forces {Fe}
with initial forces {Fo} is {6F} = {Fe} - {F }' The equilibrium
o
equations of the structure then are:

(K) {u} (27a)

Example 2

A triangular finite element (figure 8-5) has vertices I, J and


K at co-ordinates (x, y) = (0, 0), (2, 0), and (0, 1) respectively.
The material is isotropic with E = 1000 and v = 0.2* and has initial
compressive stresses Ox -10, 0y = -5, Lxy -1. Form the element
stiffness matrix and load vector.

Figure 8·5. Example 2. y Rock is isotropic with


v= 0.2
(0,1)
1 E = 1000
initial stresses are
(Jx,o = -10
(J~ ,0 - 5
= - 1

find the stiffness matrix and


int~nal forces for the triangular
finite element. (0,01 (2,0) ----x

*The dimensions for the problem are of no interest · here. Any consistent
set of units may be inferred, e.g. 0, E in MN/m2 and co-ordinates in
meters; or 0, E in p.s.f., and co-ordinates in feet. Avoid mixed
units, e.g. E in p.s.i. and co-ordinates in feet.
The Finite Element Method 319

We first form the matrix (L 1 ) , ( 16) and (16a).

(-~
n
0 1 0 0
(L 1 ) -2 0 0 0
-2 -1 0 1 2-

The stress-strain matrix is formed from (20c)

1111 278
(C) 278 1111
( o o

Thus, the stiffness matrix (K) is, by (24b)

-1 0 -2

(K)
1
4
0
1
0
-2
0
0
-1
0
1
tlll 278
0
278
1111
0
0) (-1
0\0
417 -2
0
-2
-1
1
0
0
0
0
1
0
0
2 n
0 0 2
0 2 0

695 347 -278 -208 -417 -139


347 1215 -139 -104 -208 -1111
-278 -139 278 0 0 139
giving (K)
-208 -104 0 104 208 0
-417 -208 0 208 417 0
-139 -1111 139 0 0 1111

The external forces in equilibrium with the initial compressive


stresses are, by (27)
320 The Finite Element Method

F 1 0 -2 6
x, I
F 0 -2 -1 5.5

f~~ }
y,I
F 1 0 0 -5
x,J 1
F
"2 0 0 1
y,J -0.5
F 0 0 2 -1
x,K
F y,K 0 2 -5

The nodal point displacements {u} will therefore be obtained


by solving the simultaneous equations

-5.5
5
{F } + (K){u}
e 0.5
1.
5.

CONSTANT STRAIN JOINT ELEMENT

A discontinuity can be considered as a special kind of link


between faces of blocks -- one that parts in response to tension,
slides in response to shear, and transmits any force in response to
compression. Each of these modes of deformation contributes primarily
non-elastic displacements to the rock mass. We will develop an elastic
linkage element and then, by an iterative solution procedure, constrain
the element to obey the non-elastic, non linear deformation laws appro-
priate for a discontinuity in a rock mass.
Figure 8-6 shows a four nodal point joint element as an idealiza-
tion of an actual joint. It has a small thickness (e) simulating the
irregular and variable region between the joint walls. For simplicity,
we will consider it essentially as a linear feature. We first form the
joint element stiffness in the local co-ordinates s, n.

Joint Deformation Modes


The strain displacement relationship (Lo) describes the relative
The Finite Element Method 321

actual discontinuity

y 5

~----~--------------x

finite element model


L

s=-! Figure 8·6. Joint" element.


2 II

displacement between the rock walls K, L and I, J, as a function of


nodal point displacements

{u}
s,n

where u l and vI are the total displacements of node I in the local


directions s, n (respectively parallel and perpendicular to the
original orientation of the joint walls), etc. Analogous to a
dimensionless strain quantity in the rock, we introduce joint defor-
mations defined at the center (s = 0).
322 The Finite Element Method

uK + u L u I + uJ
u
0 2 2
v + v vI + v J
K L
{E;j} (28)
2 2
vK - v v J - vI
w
L
i i

It is tempting to make {E;j} uniformly dimensionless by dividing U


o
and Vo by the joint length i; however, the definition given for
{E;j} seems preferable on physical grounds. Figure 8-7 shows how the
various modes of deformation are expressed in the joint element and
indicates the sign convention of (28) consistent with tension positive
in the rock. Relative counterclockwise rotation of the joint wall
KL is positive.
We can now identify the strain displacement matrix (L o ).

u
I
vI
I I 1
u 0 0 0 u
0 2" 2" 2" J
v
I I I J
Vo 0 0 -"2 0 0 (29)
2" 2"
uK
I I I
w 0 0 -T 0 0
T T vK

uL

vL

or {E;} . (Lo) {u} (29a)


J s,n

"Stress-Strain" Relationship for the Joints

Joint element "stresses" need to be defined, since the actual


load transfer across a rough joint may occur at point contacts. The
normal and shear stresses on the joint wall describe the total normal
and shear forces per unit of area (the thickness of the element is
unity) .
initi al co nd iti on
t///fl///////bJ of joint

VT//T/7 //77701

closed

Wffffffff~
(Vo negati ve )

open
WPJ//J'iJ/A (vo positi ve )

v~J/~/~
rota ted
! open on rig ht
partly closed
on left)

( w posi ti ve)

t'l'//~J/!//0/~
slid i ng wit h
dila t ancy

( Uo positive)
P77/T/T/T///M
slid ing with
contractancy
(uo posit i ve )

Figure 8-7. Modes of deformation of a joi nt element.


Goodman-Geological Eng.-22
324 The Finite Element Method

1
t:.a t:.a n 1(t:.F n ,K + t:.F n, L) (30)

1
and t:.T t:.T I(t:.Fs,K + t:.F s, L) , (30a)
nS

The Ildelta ll symbol (t:.) is attached to the forces and stresses to


indicate that it is the increments of force and stress caused by a
deformation {u} which we calculate. The total stress is the initi~l
s,n
stress ~quation 25) plus the stress increment. The moment (M ) of
o
the nodal point forces about the center of the joint expresses the
distribution of stress in the joint. (Counterclockwise external
moment on face KL is positive).

(30b)

Equations (30) and (30b) require a particular distribution of the


normal forces in the joint element. Inverting equations (30), (30a),
and (30b) and stipulating that FI = -F and F J = -F , we can write
L K
the nodal point forces in terms of the Ilstresses ll :

t:.F 0 0
s,I
£ 1
t:.F 0 -2
n,I I
t:.F 0 0
s,J
£ 1
t:.Fn , J 0 -2 -I
(31)
£
t:.F "2 0 0
s,K
£ 1
t:.F 0 2
n,K I
£
t:.F 2 0 0
s,L
£ 1
t:.Fn , L 0 -1;
2

or

{t:.F} (B){t:.a J } (31a)


S,n s,n
The Finite Element Method 325

Joint "stresses" and "strains" are linked through joint system


stiffness properties as follows:

(32)

or
(32a)

We observed, in Chapter 5, that ~s


and k n change with changing
stress. Therefore, the linear formulation of (32) must be coupled
with an iterative solution technique, to yield a physically acceptable
solution. One can introduce a term k
to describe the contribution
ns
to normal stress because of dilatancy (Goodman and Dubois, 1972).
As discussed in Chapter 5, in initially closed joints a shear dis-
placement in either direction will develop a joint thickening:
i.e. ~v (T) is positive when T is either negative or positive.
Th e refore k ns must be non-linear. Non linearities, including dila-
tancy, will be discussed in a later section. By omitting k from
ns
the stress-strain relationship at this stage, we simply place the
representation of dilatancy effects outside of the stiffness matrix.
The coefficient k , coupling joint rotation to the stress dis-
w

* I K

1 J
~---------------------------------~~
nPI7

l+--
~p -1

Figure 8-8. Rotational stiffness.


326 The Finite Element Method

tribution, is a new, unknown parameter. However, it can be expressed


as a function of k n if we make an assumption about the mode of
behavior. Assume that the closure at any nodal point pair across
a joint is proportional to the increment of normal force at the
corresponding nodal points. Consider the case when all the normal
force is at pair I, L (figure 8-8). The assumption demands a rotation
w = 2vo/t where v o ' as defined by (28), is the normal displacement at
the center of the joint. The moment, M , corresponding to this case
o
2
is ~~O t Since, by equation (32), Vo = ~o/kn' we conclude that
3
kw k n (t )/4. Any other case with a different stress distribution
will produce the same result. Thus we can write.

o 0

(33)

The reason to account for moment and rotation in the linkage


element is simply that the stresses and deformations in a rock mass
subjected to rotation are quite different than in a rock mass which
is not. For example, the two block systems in figure 8-9 are both in
equilibrium under the applied loading but in very different conditions.
Without rotation, all the rock blocks are in compression, and are
confined in intimate face to face contacts (figure 8-9a). In contrast
as a result of a virtual rotation, exaggerated in figure 8-9b, each
block becomes loaded by eccentric edge to face (line) contacts and
may suffer indirect tension as in a point load test .

Joint Element Stiffness Matrix

We can now write the relationship between displacements at


nodes of the joint element and the corresponding increments of
external force.

{L'JF}
s,n

or finally,
· The Finite Element Method 327

( aJ
( b)
Equilibrium without rotation
Equili brium wi th rotation

Figurf! 8-9. Effect of block rotation on stress state in jointed rock masses.

(B)(CJ)(LO){u} (34)
s,n s,n

where
(B) is defined by (31) and (31a),
(C J ) is defined by (33),
and
(Lo) is defined by (29) and (29a).

Performing the multiplication demanded by (34), the joint


element stiffness is:

ks 0 k 0 -k s 0 -k s 0
s
0 2k n 0 0 0 0 0 -2k
n
ks 0 k 0 -k s 0 -k s 0
s
0 0 0 2k 0 -2k n 0 0
(K) 1:.£ n (35)
4
s,n -k 0 -k s 0 ks 0 ks 0
s
0 0 0 -2k 0 2k 0 0
n n
0 -k s 0 k 0 ks 0
s
0 -2k 0 0 0 0 0 2k n
n
328 The Finite Element Method

It remains to rotate the stiffness matrix to the global mesh


coordinate system x, y; with reference to figure 8-6, at any nodal
point:

(36)

YJ - YI
a. = arctan

Thus we can write:

{nF} (T){nF}
s,n X,Y

and (37)

{u} (T){u}
s,n x,Y

where

cos a. sin a. 0 0 0 0 0 0
-sin a. cos a. 0 0 0 0 0 0
0 0 cos a. sin a. 0 0 0 0
0 0 -sin a. cos a. 0 0 0 0
(T)
0 0 0 0 cos a. sin a. 0 0
0 0 0 0 -sin a. cos a. 0 0
0 0 0 0 0 0 cos a. sin a.
0 0 0 0 0 0 -sin a. cos a.
(37a)
and since (T)-l = (T)T, we can rewrite (34) as:

{nF} (T)T(K )(T){u} (38)


s,n
x,Y x,Y
or
The Finite Element Method 329

{L'IF} (Kx y){U} (38a)


x,y , x,y

Initial Stresses in Joints

A joint element under initial compression or shear releases


stored energy when the equilibrating external forces are removed.
Therefore, the finite element program must associate initial stresses
with the jOints as well as with the solids . Since the joint element
is very thin, its normal and shear stresses are the same as the
stress components in the adjacent elements, referred to the joint
axis s,n.

I
t sin 20. ~
sin 20. cos a
sn,o x,o
2 2
{ao J} a
n,o sin a cos a -sin a
y,o
, s,n
0 0 0 0 T
xy,o

(39 )

or

{a 0, J} (39a)
s,n x,y

The joint stresses are given by

+ {Lw J} (39b)
s,n

Initial forces {F J} must be subtracted from the structural


0,
external load vector {F } to account for initial stresses in the
e
jOints.

(T)T(B)(T
(j ,
J){a}
0
(40)
x,y

where (T) is defined by (37a).


330 The Finite Element Method

(B) is defined by (31) and (3la)


and (T J) is defined by (39) and (39a).
0,

ASSEMBLY OF THE STRUCTURAL EQUATIONS

A structure with n nodal points produces 2 n simultaneous


equations. Denoting initial forces in the rock by F R:
0,

{F } - {F } - (F ) = (K){u} (41)
e o,J o,R

or

{F} (K) {u}


(4la)
(2n x 1) (2n x 2n)(2n x 1)

Assembly of the structural stiffness matrix (K) is exactly as in


step 3 of example 1 and is best examined through another illustration.

Example 3

Assemble the stiffness matrix and load vector to establish the


simultaneous equations for displacements in the structure of figure
8-10 (6 nodal points, 3 elements). The properties and initial stresses

A(joint)
Deformability _properties
of A: ks =500
k n=5000
L K
...---'l~v.------...

I J /'
(deformability properties of 8 and C and
Figure 8-10. Example 3. initial stresses - as in Figure 8-5)
The Finite Element Me thod 33 1

of the solid elements are the same as in example 2 (fi g u r e 8-5 ) .


Since elements Band e are identical, the rock element stiffness
has to be formulated only once if we name the corners identically fin
each element. The result was given previously (see e xampl e 2 ) .
The element stiffness for the joint, element A, is given by (35).
(Since a = 0 in the example, (T) (equatio n 37a) is just t h e identity
matrix and ( K) A
X,y
= (K)A s,n ). Let (K)N1J be a 2 x 2 matr i x of stiff-
ness terms for an element N giving the increment of for ce at no dal
point I, (F1)N, caused by x and y displacements at no dal po int
-+ N
J,(u ) Then, the 12 x 12 structural stiffness matrix for the prob-
J
lem of figure 8-10 is:

A A A A
kII kIL 0 kIJ kIK 0
A A B B A A B
k LI ( k LL + k II ) k IK kLJ ( k LK +. k I J ) 0

0
B B' e 0 B
(k KJ + k )
e e
kKI (kKK ,+ k JJ ) JK k JI
A
kA
A A
kJI kJL 0 kJK 0
JJ
A A B B e A A B e
kKI (k KL + k J1 ) (kJK + k KJ ) kKJ (kKK + k JJ + k~ ) kK I

0 0 e o ,- e e
kIJ kIK k II
~

Substituting the required stiffness terms gives (K)

250 O· -250 0 0 0 250 0 -250 0 0 0


0 5000 0 -5000 0 0 0 0 0 0 0 0
-250 0 945 347 -417 -139 -250 0 -2 8 -20 8 0 0
0 -5000 347 6215 -208 -1111 0 0 -139 - 104 0 0
0 0 -417 -208 695 0 0 0 0 34 7 - 2 78 -139
0 0 -139 -1111 0 1..215· 0 0 347 0 -2 08 -104
250 0 -250 0 0 0 250 0 -2 50 0 0 0
~

0 0 0 0 0 0 0 5000 0 -5000 0 0
-250 0 -28 -139 0 347 -250 0 945 0 - 41 7 -20 8
0 0 -20 8 -104 34 7 0 0 -5000 0 62 15 -1 39 - 11 11
0 0 0 0 -278 -2 08 0 0 - 417 -139 69 5 347
0 0 0 0 -139 -104 0 0 -208 - 1 11r 34 7 12 1 3

"
332 The Finite Element Method

Note that K i~ symmetric. Since each element stiffness matrix is


symmetric, the structural matrix will always be also.
Let F~ represent the 2 x 1 matrix of external force components at
nodal point I of element N. The initial stress contributions to the
external load vector {F} (equations 41 and 41a) are

FA
;1 I

F2 FA + FB
L I

F3 FB + F C
K J
{F}
(12 x 1) FA
F4 J

F5 FA + FB + Fe
K J K

FC
-+
F6 I

The external force contributions from initial stress in element B


were calculated in example 2. Because traangle C is obtainable from
triangle B by rotation through 180°, its external load vector is the
negative of the load vector 6f triangle B. The contributions to the
external load vector for the joint element (A) are calculated using
(40). (T) is the identity matrix (I), so that

-1 0 0 1

0 -1 +1 5
"2"
-1 0 0 1

(~ ~) flO} -
1 0 5
0 -1 -2"
{Fo,J} (I) 1 -5 -
1 0 0 -1
a -1
0 1 +!. -5
2
1 0 0 -1

1 1 -5
0 -"2

Introducing these results and changing the sign, as required by (41)


The Finite Element Method 333

the load vector is:

F -1
x,l
F
y,l -5
F -5
x,2
F
y,2 -0.5
F -4
x,3
F
y,3 4.5
F -1
x,4
F -5
y,4
F
x,5 5

F 0.5
y,5
F
x,6 6

F
y,6 5.5

Equilibrium has been disturbed by the removal of support from


the right side and the body will stretch until equilibrium is
restored. In this case, F x, 6 and F x, 5 must become zero. The dis-
placements will be found by solving the set of simultaneous equa-
tions (as in example 1), which after introduction of the fixed
boundary conditions will reduce to four equations in four unknowns.

ITERATIVE SOLUTION TO SIMULATE REAL PROPERTIES OF JOINTS

The linear equations developed for the displacements of nodal


points provide a first solution. However the indicated displace-
ments may presume tension in some joints, or excessive shear in
others; and in any case shear displacement will produce a dilatancy
tendency which has not yet been accounted for. Therefore the first
solution must be examined at each nodal point along the joints and
c ompared with real behaviour. This comparison y ields information as
to how the problem can be restarted or continued in order to produce
an entirely acceptable answer. Through an iterative process, one
334 The Finite Element Method

F
REAL BEHAVIOJR
(·constitutive lOw·)

l:Iu

Figure 8-11. Alternative iterative schemes to constrain the solution to obey the real
material behavior.

will try to compute entirely acceptable deformations and stresses


throughout the structure. These methods are similar to "predictor-
ll
corrector methods for numerical solution of differential equations
(Dorn and McCracken, 1972).

Load Transfer and Variable Stiffness Methods

Various iterative processes are illustrated in figure 8-11,


which considers a direct shear specimen with an initial shear load
(Fo) being forced by an initially compressed spring. To begin the
calculation we choose the initial stiffness k, representative of the
elastic portion of the load-deformation curve for the joint. Because
of the initial load in the spring, the spring and joint come to
equilibrium at point Rl , which is unacceptable according to our
estimate of the joint1s real load-deformation behaviour (its Ilcon_
stitutive law"). We can restart the problem with new information by
Ilaimingll the solution towards a point on the constitutive curve
near R , for example point R~. For the second run, we may re-define
l
the stiffness to k2 (variable stiffness method) calculated to pass
The Finite Element Method 335

through Ri; if we do this, the new solution will be point R2 , VS'


still unacceptable but closer to the correct answer (Roo).
To save recalculating the stiffness matrix, Zienkiewicz (1971)
suggested restarting with the same stiffness (k2 = k l ) but a new
initial load, Fa 1 (the load transfer method); in this example,
FO,l is simply F~,O + 6N l where 6N l = Ri - Rl . This yields as a
solution point R2 ,VL. The load transfer method generally requires
more iterations than the variable stiffness method to reach con-
vergence but requires fewer calculations per iteration. It is the
method used in the computer program presented in the Appendix. Other
modification paths are possible, and may be preferable in certain
instances. For example (Goodman and DubOis, 1972), if the stiffness
matrix is to be altered at each run anyway, it will be only slightly
more expensive to modify the load vector as well; then one can choose
a path such as in figure 8-11 from FCE with a slope such that the
area under FCERi is the same as the area under the actual load-
deformation curve up to RC . Such an approach would usually speed
l
convergence; however, it would not converge in all cases as shown
by Dubois (1972). All of these methods can be viewed as modifica-
tions of the Newton-Raphson scheme for non-linear functions, (see
for example Dorn and McCracken, 1972) in which the stiffness is
updated to the value of the slope of the constitutive surface in
the neighbourhood of the current solution.

Interlaced Joint Elements

The stiffness matrix, representing the coupling from one nodal


point to another, was constrrtcted from elements filling the space
between the nodal points. The " so 1 u tion fl gives the forces and
displacements at nodal pOints and it is here where modifications to
the load vector must be made to restart the problem on the next
iteration. Thus it is natural to re-define the set of joint
elements at this stage, as shown in figure 8-12. In practice it
is only necessary to store a list of lengths (£1) and orientations
(a)* to be associated with each modified joint element. In the

*If the orientation changes along a line of joint elements, the


scheme must be handled differently.
336 The Finite Element Method

Joints for formation of


stiffness matrix.
L K

I
% :J

Modified joints for


load trans fer.
a 8 7 6 5
: % I i I :

'~CDL®-L~
b

f(I n'
-1(2
n'
...K3
Figure 8-12. Modified joint elements.

example of figure 12, ti =


t l /2, t2 = (t /2 + t 2 /2) etc.
l
For computing forces from known displacements, the joint stiff-
ness properties k', and k' to be associated with a modified joint
s n
element are simply the weighted average of stiffness properties from
the two joint element halves of which it is composed. For example:

(k I ) (ks)l
s 1

tl
(k' )
s 2 -l't2- (Cks) 1 2 + (k s )2
t; )

(k~)3
1
, (Cks)2 t2
2" + Cks)3 t~) (42)
t3

(k ') (kn)l
n 1

tl t2 )
(k~)2 -l't2-(Ckn) 1 2" + (k n )2 2"

Joint Opening and Closing

In chapter 5 it was presumed a joint can not sustain tension and


furthermore, that it becomes thinner under a compressive normal load,
The Finite Element Method 337

until a maximum closure V (figure 8-13). The following discussion


mc
illustrates how the law of normal deformation (equation 8 of chapter
5) can be obeyed by means of a load transfer process. Substituting
t = 0 1 A = 11 and; = F
nlo
It' in equation 8 of chapter 5 gives

F (43)
nlo

where: Fis the initial external force at a nodal point; ~v is


nlo
the difference of normal displacements between the individuals of a
nodal point pair caused by an increment of normal force (F n - F n,o )
and Vm is the maximum closure beginning from initial load Fn,o .
For the load transfer operation assume (43) applies both for loading
and for unload,ing.
Let ~v be the joint opening (positive displacement) on
releasing the initial stress (on = 00) to the seating pressure

Tension

(j _ Fn
-7

- - Vmc

Vm

(opening)

Figure 8-13. Relationship between normal stress and normal displacement


at a nodal point pair.
338 The Finite Element Method

~ used to measure V From (43) ~v V - (F /F)V


°Vn mc m
=
n,o n m
m Since V
me ~v(~) - Vm (figure 8-13):*°

-V
V mc ~
(44)
m
°o

The computation is initiated with a unit normal stiffness k


3 n,o
(force/length) derived by differentiating (43) with respect to ~v:

2
-0
k
n,o
°o
V
o
(44a)
m ~Vmc

where °0
is the initial normal stress in the joint element; this
value of normal stiffness produces a solution with displacements
Va and Vb at nodal points a and b along a modified joint element **
then:

(45)

(in general, there will also be a tendency for dilatancy ~V(T),

as will be discussed later). From the initial stress °0 , reference


pressure and maximum closure Vmc of the two joint element halves
~,

forming the modified element, we find the normal stiffness (k')


n
(stress/length) using (44a) and (42). If both joints comprising a
modified element have the same initial stress and maximum closure,

F
°o ~ (46)
Vm V ,Q,'
m

*Assuming unit joint thickness. ~ and °0 are negative since they


are compressive. Vm is a negative displacement. k n and Vmc are
defined to be positive.
** In the modified element, a is K or L of a joint element while b
is I or J of a joint element (figure 8-12).
The Finite Element Method 339

where £' is the length of the modified element and V is given by


m
(44) .
Then, the external force at node a of the pair a, b is:

F £ ' k ' (6. v ) + F (46a)


n,l n l n,o

Thus, as in figure 8-14 the first iteration produces point Rl


(Fn,l' 6.v l ). Only in a rare instance will Rl be precisely on the
constitutive curve (43).
Joint opening (6.v positive) will be considered first. In
l
figure 8-14, 6.v is positive, and Rl is above the curve. A point on
l
the constitutive curve with 6.v = 6.v l defines point Ri (compare with
figure 8-11) and the distance Ri - Rl determines the initial load
for the second iteration, (F n,o )2 as follows (figure 8-14):

(F n,o )1 + 6.N
l

Initiol slope = k~ f
:_0_
0:: g'
Vm RI(Fnl,~VI)
R2 (Fn21 Ll v2)
Vm R3R4

Figure 8-14. I terative process for joint opening (.6 v positive).


Goodman- Geo logica l Eng .-2 3
340 The Finite Element Method

where (47)

F F
n,o n,l

For the (i + l)th iteration

F . ~'k'~V. + (F ).
n,l n 1 n,o 1

and (47a)

where

~
~v.
1
~N. F F .
1 ~v. n,o n,l
1 +

and (47b)

~v.
1

The process must be repeated until ~N. is smaller than a


1
satisfactorily small number E in all elements, and other non-linear
constraints to be discussed have also been satisfied.
Combining equations (47a), and (47b) and introducing (45) gives
a load transfer formula for joint opening:

~v.V
( (Vm
1 m
-~v.)
1
+ (V - IIV)
m 1; (48)

The iterations are the same whether Ri (figure 8-14) is above or


below the constitutive curve; in the latter case, application of (48)
will produce a positive value of ~N.. Occasionally, iteration will
1
yield an oscillating convergent series of ~Ni values.
The Finite Element Method 341

Joint closing is handled similarly, except that t h e po int R~


must be guided by the force rather than the displacement computed in
the previous iteration (figure 8-15). For t h is purpose we rewrite
the constitutive law (43) as

vm(Fn - Fn , 0)
8.v (49)
F
n

The results of the first iteration now show 8.V l negat ive, with point
Rl = 8.V l , Fn,l' The initial load for the second iteration is:

(F n,o )1 + 8.N l

where

I
k Q., (50)
n

For the (i + l)th iteration

Vm c -----~~I
Vm
---j---j---------~~~~~-------6v

F: - F
l!Jv=V m ( n Fnn,o
)

Figure 8-1 5 . Iterative pror.ess for joint closing (6 v negative).


342 The Finite Element Method

F . £'k'~v. + (F ).
n,l n 1 n,o 1

and

(F ). + (~N.) ( 50a)
n,o 1 1

where

vm(F n,l. -
k' .Q,'
F . n
0,1

Combining equations (50a) and introducing (45) gives a load transfer


recursion formula for joint closing:

1
= kn £
, '( 2~v.
1
- V +
m ~v. 1
V 2
+ (F m )./(k'£')
n,o 1 n
)
(51)

As in the case of joint opening, iterations are required whether


Ri is to the left or to the right of the constitutive curve; in the
case depicted in figure 8-15 with R. to the left, ~N is negative,
1
whereas if Ri is to the right ~Ni will be positive. Sev~ral examples
will be discussed later.

Joint Shearing and Dilatancy

Joint Shearing can be treated in the same fashion as joint


opening; the limiting shear stress criterion to be imposed on the
shear-deformation between blocks is analogous to the "no tension"
criterion imposed on the normal stress-opening curve. Since we lack
a universal model describing the shear deformation-shear stress
behaviour of joints, a simple constitutive law will be assumed
(figure 8-16). The initial stress, LO' and initial shear stiffness,
ks define the elastic region, whose limiting stress Lp depends upon
a according to a criterion of peak shear strength:

(52)
The Finite Element Method 343

Tp = fl (O')
Tr = f2 (o-)

I n m

Figure 8-16. Constitutive law for shear deformation.

If Lp is exceeded, the strength falls, attaining a residual value,


L , when the displacement u r has been attained.
r

(53)

Unfortunately, we do not yet know enough about the variation of peak


and residual displacements (up' u r ) with normal stress 0. As new
information becomes accepted, it can supplant the model assumed here
which is: for L positive (right lateral shear)
* '

u , are defined as positive quantities.


p
344 The Finite Element Method

L - L
P 0
k
s
(54)
u
r
M > 1.

For L negative,

(54a)

Let

(54b)

be the relative shear displacement in a modified joint element; then


the constitutive relationship (figure 8-16) may be expressed formally
as follows:

L - L
r
II L -L
p + u2 - u
(~u - up (_) ) ; u r (_) < ~u < u
p(-)
p r

III L = k ~u + L U < ~u < U


S 0' p(-) p(+)

L - L
r
IV L = L + E (~u - up (+)) ; U < ~u < u (+)
p up - U
r
p(+) r

V L = L ~u > u (55)
r' r(+)

As before, the load transfer is in terms of forces in the modi-


The Finite Element Method 345

fied joint elements. The computation, begun with a stiffness ks in


each joint element, gives ~ul = u a - u b (figure 8-12). The corres-
ponding shear sorce at nodal point a is

F (56)
s,l

where k's is given by (42) and (F s,o )1 = T 0 £' (compare with (46a».
Then, as for joint opening, the initial force at node a for iteration
2 is

(F s,o ) 1 + ~Sl

with (57)

F
s,l

where Ll is given by (55) with ~u = ~ul.


Similarly, to begin the i + Ith iteration,

£'k' ~
F
S,l
. s u i + (F s,o ).1

and

(F ). + ~S. (58)
S,O 1 1

with

L.£' F .
1 S,l

Combining equations gives:

L.£' (59)
1

where T.
1
is given by (55) which in turn depends upon the choice of
fl and f2 (52) and (53). Any consistent specific experimental or
346 The Finite Element Method

Figure 8-17. Assumed variation


of peak and residual shear strength t
and peak dilatancy (v) with normal
stress; based on Ladanyi and
Archambault (1970).

To or = qu (assumed) cr

qu = urlcorlfirled compressive
strength

empi;ical results can be used to define (f l ) and (f 2 ). As an illus-


tration we will input the formulas given by Ladanyi and Archambault
(1970), chapter 5, equations (14), (15), and (16), (figure 8-17).
Unfortunately we know little about the variation of residual
shear strength, T r , with o. We will input f2 in a consistent manner
as follows. At high normal stresses, rock becomes plastic, i.e.
exhibits a ratio T r IT p = 1. We will presume therefore that the ratio
T IT increases from B , (0 < B < 1) at 0 = 0 * to 1 at 0 = q .
r p 0 o - u

Tr Tp (Bo 1:uB~o
+

and (60)
T .
p'

*This value will have to be obtained in general by extrapolating


data to zero.
The Finite Element Method 347

Bo can be left as an input parameter for a computer program to permit


one to contrast the behaviour of brittle and plastic joints.

Dilatancy must also be introduced into the analysis. Dilatancy


describes the normal displacement (joint thickening) anv(T)/aU
caused by shear. Let v= nV(Tp)/U p (the secant dilatancy rate).
Then as an approximation, the normal displacement caused by dilatancy
at a shear displacement nu is:

< nu <

and (61)

nV.(T) nu > u r (+)


1

or nu < u r (_)

The absolute value function and the minus sign insure that the joint
thickens regardless of the sign of the shear stress, as discussed in
chapter 5.
The variation of v with a was given by equation l6b of chapter 5.
Substituting this equation in (61) with aT = qu * and considering
stresses in the modified joint element gives for the ith iteration:

CFn)i_l ) 4
, - 1 tan i
( ~ q 0
u

for u r (_) < nu < u r (+)

*As noted previously, qu is here a negative quantity, since tension


is positive.
348 The Finite Element Method

and (62)

for flu > u r (+)


or flu < u r (_)

At the end of the ith iteration, we will know the shear dis-
placement flu. at each nodal pOint pair. Inserting flU in (62) for
1 i
each modified joint element, we determine flV.(T).
1
If all dilatancy is prevented by the adjacent elements, there
must be external compressive forces applied to joint nodes a and b
equal in magnitude to the dilatancy multiplied by the joint stiff-
ness. Redefining the jOint stiffness as the slope of the compression
curve (43) evaluated at the previous normal stress (F . l/~')' the
n,l-
increment in initial normal stress in the joint due to dilatancy
calculated for the ith iteration is

flV.(T)
1
(fla
n,o
).
1
= ["V
mc
(62a)

To start the i + 1 st iteration, the initial stress incremented by


(fla
n, 0)'1 produces external forces on the neighboring elements, which
in fact deform, thereby automatically relaxing the initial assumption
that dilatancy is prevented.
For the first iteration we must use (F ). 1 F
n 1- n,o
Updating the Loads

The final 'step' in the load transfer procedure is to rotate


(Fn,O)i+l and (Fs,O)i+l to global coordinates and update the load
vector. With the sign convention used in the modified joint ele-
ments, (F ). and (F ). are external forces at nodal point a (figure
n 1 s 1
8-12). We will update the load vector as follows. If a nodal point
is not in a joint, the load at that nodal point does not change. At
every nodal point pair a, b belonging to a modified joint element,
the new load vector terms are:
The Finite Element Method 349

F cos ex. sin ex. 0 0


x,a

F
y,a cos ex. 0 0
(63)
F 0 0 cos ex. sin ex.
x,b

F 0 0 -sin ex. cos ex.


y,b i + 1

{F}i+l is then multiplied by the inverted stiffness matrix (k) - l


to yield the new estimate of displacement {u}i+l. Table 8 - 1 summarizes
the steps in the load transfer scheme, as discussed here. *
SOURCES OF EXTERNAL LOAD
The contributions of residual stress to the load vector have
been explored . Here we will consider additional forces from water
pressure distributions, gravity, pseudo-static accele ra ti o ns (inertia
forces), and active or passive supports.

Water Forces

Water forces tend to change as a result of the joint deforma-


tions they cause. Therefore, any complete analysis of the action of
water becomes a problem in coupled stress and flow, as discussed by
Rodatz and Wittke (1972), Noorishad et ale ( 1972), Gale (1975), and
others. This class of problems, which will not be treate d here, can
be set up as follows. One assumes an initial water press u re distri-
bution, which generates force contribution at the nodal points.
Iterative solution by load transfer as discussed in this chapter

*Certain economies are realizable by following a somewhat mo d ified


scheme for load transfer. Store the displ ac ements {u i }. Use terms
6N i and 6S i , rotated to global co-ordinates, for each mod ifi ed joint
element to build a force increment vector {6F . }, which when multi-
1
plied by (k)-l gives displacement increments for all no d al po in ts.
(The force increments at nodal points inside solid eleme n ts are zero).
The incremental displacements are cumulated and added to the dis-
placements of the first increment. This modified procedure replaces
many products by zero but requires summing displacements.
TABLE 8-1
Summary of Steps in Finite Element Analysis
of Jointed Rock Masses by the Load Transfer Method

i = 1 (Parentheses identify relevant equation numbers)

1. Read input: geometric and material properties; initial stresses;


accelerations; pore pressures; and support loads.

2. Form solid element stiffness matrix for each type, orientation,


and shape of solid element. (24a)

3. Form joint element stiffness matrix for each type length and
orientation of joint element. (35) (38)

4. Assemble structural stiffness matrix (K) (see example 3).


(Actually this step is performed simultaneously with 2 and ~)

5. Assemble residual stress contributions to the load vector (27)


and (40). (This step is also done simultaneously with 2 and 3).
Change sign and store in load vector (41).

6. * Add external forces from other sources: water pressures; weight;


active and passive supports. Total load vector = {F} 1..
7. Invert the structural stiffness matrix (K). For small computers
this may be done outside the rest of the program as it only
needs to be done once. Store (K)-l

8. * Determine displacements by matrix multiplication.


{u} = (K)-l {F}.
1

9. Form modified joint element stiffnesses and initial forces (42)


(46) and relative displacements 6V i and 6U (45), (62) and (54b).
i
10. Determine normal force {Fn}i (47a) and shear force {Fs}i (58) in
each modified joint element.

11. Find 6N and 6S in each element (47b) or (50a), and (58). If


~6N + ~6S > E, where E is some small number, update Fn , 0 and
F by (51) and (59). Rotate to global coordinates and update
s,o
{F}i·
12. i = i + 1. Go to step 8.

*For incremental loading, divide total load vector into small


increments and cumulate their effect.
The Finite Element Method 351

defines the change of volume of each modified joint element.

livol livol
8 (64)
vol £'e

where e is the initial thickness of the jOint. The change in water


pressure in the modified element is therefore:

livol ( 64a)
C £'e
w

where Cw is the compressibility of water.


The result of the first iteration will thus produce a new water
pressur~ distribution which can be expected to initiate flow through
the network of joints. We allow flow to occur until a new pressure
distribution has been established, calculate the new water pressures
and begin the stress problem anew. The problem is unfortunately
quite path dependent .
. In each triangular rock element, water forces are input as
applied external forces at the nodal points in the direction opposite
to the hydraulic gradient. If as a simplification we assume the
pressure gradient is constant over the element, it follows mathe-
matically (Zienkiewicz (1971) that the water forces distribute equally
to each node. . *:
Taking the global y axis positive upwards glves

dh
F
x,I
- dX

F y,I 1 - dh
ay-
F dh
ax
~ext}water
x,J ywa
a
"3 -3-
F dh
1 - ay- (65)
y,J
dh
F
x,K --ay
dh
1 - -ay
Fy,K

* dh/dY is positive for downward flow; dh/dx is positive for flow to


the left.
352 The Finite Element Method

Figure 8-18. The equipotential lines of a 60° saturated slope


(after Han, 1972).

A B C D E
dh/ch -0.79 -0.06 -0.04 -0.20 -0.1,3
ohley 0. 00 .0.34 - 0.13 .0.07 .0.58

160 180 200 220 21,0 260

For example figure 8-18 shows lines of equal potential for a


0
60 saturated slope, obtained by Han (1972) using a conducting paper
method. The gradients for elements centered at five points have
been calculated from the diagram and are tabulated on the figure.
These values will be input in the respective elements and the nodal
point forces computed internally by (65). (Above the water table,
we assign y w = 0.)

Weight Forces

The initial stresses input in the analysis may reflect the action
of self weight forces on the region under investigation. But the
residual stress input can not duplicate the action of gravity because
"loading" by an initial stress and the loading of gravity are
fundamentally different. The residual stress load is like that of a
precompressed spring; it can not continuously follow the deformations
it causes in its neighbours. Gravity, on the contrary, produces an
inertia force that pursues the deforming body. Therefore, to simu-
late the gravity action in a beam we require additional forces input
The Finite Element Method 353

as external forces at the nodal pOints; (we should also reduce the
initial stresses accordingly so that gravity is not counted twice).
Since gravity produces a force distributed over each unit of mass,
it can be treated like the seepage forces just considered. Denoting
by y the weight per unit volume of the rock (total, not buoyant),
the required applied external forces in an element are:

F 0
x, I

F
y,I -1

F 0
x,J
~ (66)
3
F
y,J -1

F 0
x,K

F -1
y,K

(As before, we assume the global y axis is positive upwards).

Pseudo Static Acceleration

Pseudo static acceleration can be included for purposes of a


limit equilibrium analYSis. If a body is accelerated kg in a direc-
tion 180 + a degrees from the positive x axis (measured counter-
clockwise from x), it will experience inertia forces in the direction
a; these forces can be input with gravity by rewriting (66) as
follows:

F k cos a.
x, I

F
y,I -1 + k sin a

F k cos a
x,J
~ (66a)
3
F
y,J -1 + k sin a.

F k cos a.
x,K

F
y,K -1 + k sin a
354 The Finite Element Method

Rock Bolts

Tensioned rock bolts provide a pair of forces compressing the


rock at the bearing plate and anchor ends. One must layout the con-
figuration of nodal points with the locations of rock bolts in mind.
The forces are simply added as external loads parallel to the bolt
axes at the nodal points in question. Rock bolts produce two addi-
tional effects, however. First, the steel stiffens the rock. This
can be included by increasing the appropriate stiffness terms for the
nodal points along the line of the bolt. It may be more satisfactory
to input one-dimensional bar elements along the line of the bolt
(see example 1); these should be constrained to yield at the appro-
priate load through load transfer. Secondly, the bolts act as shear
keys in crossing each joint; since the bolt is confined by the rock,
the rock must crush around the bolt to permit it to shear. The bolt
action can be simulated by supplementing the peak and residual shear
strength, of the modified joints containing the rock bolt, by an
amount equal to the shear strength of the bolt, plus an increase in
parameter as in Ladanyi's peak shear strength criterion.

EXAMPLE PROBLEMS

Example 4-Joint Closing

In figure 8-19, a block under an initial stress of 5 MN/m2 com-


pression is next to a joint with initial compression 1 MN/m2, held
by a constraint (not shown). When the constraint is removed, the
momentary disequilibrium destresses the block and compresses the
joint to restore equilibrium. The speed of convergence depends
upon the initial stress in the joint.
Putting eq (44) in (43) and substituting F 0 £'~ and ° =~,

an (-Vmc~~ ~v)~ +~. The normal deformation :~rve correspon:ing


to V
mc
= 0.05 and ~ = -0.1 passes through the points (F ,
n
~v) in
Table 8-2.
As shown in the graphical solution (figure 8-19) about 5 itera-
tions are required for convergence for the given data, with 00 =
-1 MN/m2 in the joint, whereas when a is -0.5 and -0.2 MN/m2 con-
o
The Finite Element Method 355

(0,1) (1,1) Figure 8-1 9. Examp le 4-jo int closing.


5 67

UX,o= -I

2 3
(0,0) (1,0)

Propert i es '
Rock : E = 100 MN/m 2, 11 :0
Joint: when t = - 0.1, Vmc =0.05m
Relative Displacement 2 ~

If) 0 If) 0 k n,o = ~


t V = 200 MN('1m 3
- _ 0 0 mc
o 0 0 0
9 9 '2 0
-0.06 !Jv(m) 0

I nitiol stress in joint -I

-2

-3

-4

SOlution

I nitio I sf ress In block -5

vergence requires 12 and 35 iterations, respectively. In s u c h a


case, convergence can be accelerate d .

Example 4a-Accelerating the Convergence

Had we carried out example 4 with an initial stress of -0.2 MN/m2,


Good man-G eo l og i cal Eng -24
356 The Finite Element Method

TABLE 8-2

!::'v (j F
x n

0 -.100
-.010 -.125
-.020 -.167
-.030 -.25
-.035 -.333
-.040 -.500
-.042 -.625
-.045 -1.0
-.046 -1.25
-.048 -2.5
-.049 -5.0

35 iterations would have been required for convergence (figure 8-20).


It is likely that the requested number of iterations would fall short.
On restarting to continue the computations, several ways to accelerate
the convergence are possible, as in figure 8-20. Say the first com-
putations with initial joint stress at A terminated after 2 itera-
tions, yielding point R2 , and associated points C (=R~) and E (=F o ,3)'
Three alternative accelerating schemes are shown.
(a) A Newton Raphson correction restarts from C with stiffness
redefined to the tangent to the compression curve at C. The
initial stress in the rock must then be changed, for restart,
from B to D or the 'solution' will shift. The displacements
will have to be stored from the first run and added cumula-
tively on each restart. Convergence will be achieved in 6
more iterations.
(b) Another approach, the dashed lines, restarts at E with the
stiffness redefined to the slope of the compression curve
at C. This produces an over correction but convergence
is reached in 7 more iterations.
(c) A third approach is given by the dotted lines. The restart
point is the initial starting point A, and the stiffness
is redefined to that corresponding to line AC. Convergence
The Finite Element Method 357

flv
--~--~~----~--~~~----~----~~----~----~r-----~O

-I

-2
Solution
Method
- - unmodified
- - - - Newton Rophson -3
- -modified (I)
........... modi fied (2)

-4

Figure 8-20. Example 4A-accelerating -5


the convergence.

is achieved in an additional 16 iterations. The first run


writes the new joint stiffnesses on tape which is read on
restarting.
With reference to figure 8-15, the unit normal stiffness in a closing
joint upon restart after i iterations is

k .
I F . - Fn,o
n,l F
n,l
.
(2)
n,l t !:.V.
1
-W-m

Example 5-Joint Opening

Example 5 is a case of joint decompression (opening) which


eliminates an initial block tension. Figure 8-21 shows a graphical
solution with convergence in 22 iterations. Restarting after the
second iteration with redefined normal stiffness according to modi-
358 The Finite Element Method

fied methods 1 and 2 accelerates the convergence to 9 and 18 itera-


tions respectively. The computer output for the first 5 iterations
gives the results of Table 8-3.

TABLE 8-3

iteration (J in solid ~N at Nodes 2 and 6 u at nodes 2 and 6


x x

Start 1.0 o o
1 0.667 -1.52 -0.00333
2 0.497 -1.16 -0.00503
3 0.368 -0.93 -0.00631
4 0.264 -0.76 -0.00736
5 0.179 -0.64 -0.00821

The modified unit normal stiffness corresponding to modified


method 2 (figure 8-21, dotted lines) will be calculated for the case

of joint opening. With reference to figure 8-14, the stiffness
(stress/displacement) for accelerated restart after i iterations in
an opening joint is:

(F . - F ) F
1 n,l n,o n,o
k . (68)
n,l £ ~v

Examples 6 ~nd 7-Sliding


These examples show how a shear failure is indicated by diverging
output. (Unfortunately, numerical difficulties can sometimes produce
divergence in a stable case so that 'failure' can only be suggested by
the output, not confirmed by it). A block under an initial vertical
stress is pushed by a 'following' force past two joints, whose peak
strength is insufficiently large. In example 6A (figure 8-22) the
joints exhibit peak-residual behaviour (Bo = 0.333) while in
example 6B (figure 8-23) the joints are plastic (Bo = 1.0). The
addition of a 15° dilatancy angle to the joints, example 7, (figure
8-24) brings stability in both cases. The dilatancy raises the peak
shear strength for the first iteration (from 0.66 to 0.83) but this
is still short of the applied stress (1.0). However, the shear
The Finite Element Method 359

Joint Opening
Block and joint of Example 4
0"0=-2 .0
wit h init i 01 tension in block
Co m press ion

Convergence
Converges in 22 iterations. 2.0
With accelerator applied after
2 iterations:
Relative
- - Mod. Method 1-10 iterations
-0.01 -0.02
........ Mod. Method 2-11 iterations

L------r-r+H~~~--~------~----~------~--------~0

/
I
/ /
/ / -2.0
/ I
/
/
/
/
/ Block
/
/ -4.0
/
/
B
Figure 8-21. Example 5.

displacement on the first iteration raises the normal stress which


further increases Lp above 1.0. Table 8-4 contrasts the convergent
solution of example 7, which is stable, with the divergence seen in
example 6A, in which f ailure occurs.
360 The Finite Element Method

Figure 8-22. Example 6A. ~=-0.520


(0 I)-w "'# 7 ®l 8
'., -.. (II)

5~6
0 .5 N '

CD 4 @

N O"y,o=-0.520
0.5 .. I ® 2

~\ / "A- 3 ®~ 9
(O,~ __ (1,0)
O"y,o - 0 .520
Rock properties as in Example 4
J oint properties:
V =0.05 ~ = 0.1
mc
Ks = 20 MN/m 3 Bo= 0.333

4>fL = 30° LO = 0

Fx
{MN
1.01----------iF------ii~---~----__.,.__-

0.8

06

0.4

0.2

Diverg es - Unstable

TABLE 8-4
Output Displacements in Examples SA and 7

iteration
x at node 4
u (m)
example 6A example 7
1 0.0500 0.0500
2 0.0710 0.0587
3 0.0963 0.0500
4 0.1271 0.0500
The Finite Element Method 36 1

- - -Accelerated restart
after 3 iterations
Fx
(MN)
f.OI---"-------~-___:~-____:>.~--------

Diverges - Unstable

Figure 8-23. Example 6B-shear failure on non-dilatant joints.

2.0

Figure 8-24. Example 7-shear on dilatant joints.


362 The Finite Element Method

Accelerated restart is performed as shown by the dotted path in figure


8-23. The revised unit shear stiffness for restarting after i
iterations is

k . (69)
S,l

Example 8-Rotation of a Block

A square block between two horizontal joints rotates in response


to a clockwise moment (figure 8-25). The first iteration produces an
overclosing of the joints. The solution after 10 iterations shows
normal displacements within the limit; as the block deforms to meet
the maximum closure restraint, the stress state changes from biaxial
to approximately uniaxial, directed along the block diagonal. A
larger loading of the same style would produce tension in the solid.

~\\\\\\\\\\~~ Figure 8-25. Example 8-effect of joint closing


on stresses in a rotating block.
~r---------.....,

I Initial state

~-------~~

~!77111171/////7/~ Displacements and joint


thickness exaggerated
~;;.:r'00~V\\\~ 10 times ~\\\\\\\\\\\\\\\~
I
~

----, > -1.0 <:


--"r--
I
-- -------- -,
/ I
I I I

~
/ stress magn. scale
I I I

X
/ I I I Results after
I ~
I 10 th itera t ion
I I Results after I
/ I 1st iteration I
I I I
l.
-- -- -_ I
/
/
--- - I
--../~
I

,c;;mlIIIlmll77l1,},.
The Finite Element Method 363

EXAMPLE 9 EXAMPLE 10
Gravity "Turn-on" Gravity -Turn-on" plus 0.5 9
Horizon tally

Mesh 0 10 20 30 40 m
I I I I I
Displacement 0 0.2 0 .4 0.6 0.8 m

Figure 8-26. T he development of toppling.

Shearing and Overturning of Rock Slopes-Examples 9, 10, 11, and 12

Figure 8-26 shows intermediate states in an unstable case invol-


ving overturning of a slab on an overhanging slope. The "turn-on "
of gravity, example 9, produces an opening af the disco nt i n uity at
the top of the slope, and a closing at the toe; however the rate of
overturning is very slow without an accelerated restart. In example
10, the rate of divergence is much faster because a hor iz ontal
acceleration has been superimposed on gravity. Figure 8- 27 s hows
the importance of accurately reproducing the path of actual load ing
in these problems involving gravity. In example llA, gr a vity h as
been "turned-on" and the downslope movement of the wedge is con fuse d
with the lowering of the surface as gravity is applied. A better way
to perform this analysis (example 11), is to restrain t he free
slope for the first run in which gravity is applied; th en int roduce
the stresses corresponding to the results of the first r u n a s i nitial
st re sses while a second run is made with the slope re s traints removed.
The only deformations remaining will be those associated with the
downslope sliding of the block (presuming the stresses i n d i cate d in
the output to the first run are in equilibrium and accept ab le in all
elements). In example 12 (figure 8-28) an unstable c ase is i nd i ca te d
364 The Finite Element Method

MESH
89

Displacements 200 X

12 (100,0) Mesh 0 20 40 60 80m


® 7 I I I I I
»-----~---=-.I. 4 Osplacement 0 0 .1 0 .2 0.3 0.4 m
(0, -20) (100,-20)

DEFORM ED SHAPE AFT ER 10 ITERATIONS

EXAMPLE "A EXAMPLE II


Gravity Turn-on Analysis Initial Stresses Equal to Gravity-
induced Stresses Plus Gravity
Acceleration

en
~

~ 0..04
E
~
ci
~
a.
0.02
en Converging
(5
N
·c
~ OL-~-~-L-~-~-~~-~-~~----
o 2 4 6 8 10 Iteration

Figure 8-27. Effect of loading path in a gravity-loaded wedge.


The Finite Element Method 365

Figure 8-28. Failure of a slope indicated by diverging output.

DEFORMED SHAPE AFTER 10 ITERATIONS

Mesh 0
I
10
I
20 30 40m
I I I
!!~~ 2l
1.0
o a. 0
______
I I I I
-----
I I I
O;,.,g;ng
I I I
Displacement 0 I 234m reO 2 4 6 8 10 It eration

EXAMPLE II A EXAMPLE 12
Gravity Turn- on Gravity Turn-on plus 0 .59 Horizontally
"stable" ?

by the output divergence when a horizontal acceleration of O.5g is


superimposed on the gravity acceleration.

Example 13-Study of Underground Excavation

Figure 8-29 shows a mesh used to investigate progressive rela-


xation and failure of rock around an underground power station. To
model the layers of shale, coal, sandstone, and siltstone, 288
elements and 329 nodal points were used in a load transfer program. *
Since every bedding plane could not be represented, the stratigraphy
was simplified by division into a manageable number of units. Rows
of joint elements were placed between each of these units. A cal-
culation provided anisotropic properties for each layer which took

*This analysis was performed in collaboration with J. Dubois and


T. A. Lang.
366

-I-- --+- 2
-f-J...;-! S 10
--+-c -f- -f- ""i- -f-
+
+-

+ -I- /J
-r----+- ........ -+-&'
~~~
""'f..c .,
~ D

/0 ·

>-[- 219

+ 24 1

+2H
+ 7S~

+ ~Sl
+
"". +
-.fO
+164 +
·1tJ

+ ,u
-$tJ +m + 271 !I()

Figure 8-29. A finite + 276 U7


+m
-element study of an -- - +
underground
excavation.

into account the number of individual bedding planes it might contain.


The mesh at first included the first stage excavation as an initial
opening. When the initial stress believed to exist in the rock was
input, the walls and roof of the stage I excavation deformed and the
stresses changed; some elements were over-stressed or placed in
significant tension. Judgement was applied as to whether or not such
elements might "fall out" and if it was thought they could, additional
computations were run with lower stiffness properties prescribed
for these elements. Eventually the second stage of excavation was
effected, elements were "removed" or "modified" in properties accor-
ding to the severity of the stresses and subsequent excavation stages
The Finite Element Method 367

applied. Figure 8-29 shows the results at an intermediate stage of


excavation.

INCREMENTAL LOADING

The opening and closing, shear and dilatancy of discontinuities


introduce a large degree of non-linearity into structural calcula-
tions. If there are numerous jOints, and if loading creates large
changes from a stable initial condition, the iterative process may
not converge. It is a good practice, in such cases, to use an
incremental solution, in which the loading is divided into small steps
and applied cumulatively. For a gravity loaded problem, for example,
an initial increment of perhaps ten percent of the gravity load
vector is applied and iterations pursued by load transfer or by
variable stiffness methods until convergence is approached. Then,
the next increment of load is applied by increasing the acceleration
of gravity by another ten percent, and the convergent displacements
and stresses are accumulated. The process is repeated until the
whole load has been applied.
For an initial stress problem, where the initial equilibrium
state of stress is to be disturbed by an excavation, the normal and
shear loads on the excavated boundary must be zeroed gradually. This
can be done by adding normal and shear forces to these boundaries
which are directed opposite to the initial forces and are of say
ten percent of their magnitude. The excavation will then slowly
approach the final configuration as the unload increments are accu-
mulated. Thus, incremental loading can be used for both loading and
unloading problems to improve convergence.
In initial stress problems, one must be careful to insure that
the input stresses are in fact in equilibrium. With heterogeneous
or anisotropic properties, this may not be easy. It would be for-
tuitous if measured residual stresses were all compatible with the
geological structure and input properties. Thus, an initial cycle
of iterations may be warranted, using the assumed initial stresses
in a mesh without a free boundary, and without any load or unload
increments. When this problem is computed, there will be displace-
ments inside the mesh and changes in the initial stresses. These
368 The Finite Element Method

adjusted stresses can be used as improved estimates of the initial


stress state, since they will be in equilibrium.

RIGID BLOCK ANALYSIS

When there are a large number of closely spaced joints in hard


rock, the joint deformations so overshadow the block deformations
that the blocks can be considered rigid. Economies in computer
storage are thus possible in which case larger systems of blocks can
be accommodated. Calculation of the deformations of a rigid block
system begins with input of incremental forces in the joints. These
are transformed to increments of force and moment at the centroid
of each block. For a short time step, the resulting accelerations
of each block are integrated to give new positions and orientations
for the block centroids. As a result of these block motions, block
sides and corners are displaced, deforming the joints and causing
new increments of joint forces. This starts the second cycle of
computation, etc. By integrating with small time steps, large defor-
mations can be accumulated. Cundall (1971 and 1974), Burman (1972),
and Byrne (1974) have contributed to the development of rigid block
analysis. Cundall programmed the solution with interactive computer
graphics with simplified input and output. The subject holds great
potential, especially for kinematic study of large block systems
under either static or dynamic loads.
appendix

ONE
an illustrative
finite element prog am

PURPOSE AND SCOPE OF THE COMPUTER PROGRAM

This appendix presents a computer program based upon the material


of Chapter 8. The program was written by Dr. Christopher St. John*
in collaboration with the author. The sole motivation for preparing
this program was to enable an interested reader to follow the coding
of equations for analysis of jointed rocks. Thus, the program has
been kept small, mainly by omitting obvious refinements and generali-
zations. However, even a more sophisticated and more general program
would soon become obsolete in this rapidly moving field. If you
have an acquaintance with the Fortran language, the processes and
algorithms presented will help you to understand the principles
better.
As presently structured and dimensioned, the program JETTY,
listed at the end of this appendix, calls for a digital computer
having a 21,000 word core memory and will solve problems with up to
25 nodal points, 23 elements, and 20 different types of materials
or joints.
A number of improvements and enlargements beyond the scope of
this program will be desirable when using the finite element method
in practise. Constant strain triangles were used here to simplify
the representation of rock. Better and more convenient elements are

* Lecturer, Royal School of Mines, Imperial College, London


370 Appendix 1

now available. In addition, special elements can be introduced to


represent beams, tunnel liners (shells), rock bolts, and cables.
A Berkeley computer library matrix inversion routine was called to
solve the structural equations. For use elsewhere, this can be
replaced with the CDC routine MATRIX, or other direct solution schemes.
However, the stiffness matrices used here are symmetric; furthermore
they can be narrow banded, i.e. numbered so that the relatively
few non-zero terms cluster about the main diagonal of the structural
stiffness matrix. Therefore, when solving larger problems, one will
be able to economize greatly by using symmetric, narrow banded
equation solvers such as BANSOL of Professor E. Wilson.* Also, to
extend the size of a program beyond that which can be stored simul-
taneously in core, it is possible to utilize supplementary storage
with a block solution procedure, as done in BANSOL. Plotting routines
prove invaluable in debugging large meshes and in interpreting the
output displacements and stresses.
With respect to problems with discontinuities, it may sometimes
be desirable to introduce algorithms other than the load transfer
procedures discussed in Chapter 8 and incorporated in the subroutine
JSTR. In particular, when the initial normal stress in joint elements
is either very low or very high, the initial normal stiffness may
prescribe a structural stiffness which requires too many iterations
to converge. For such problems, a variable stiffness approach is
more appropriate. The load transfer method was used because the
stiffness matrix has to be inverted only once; therefore, joint
non-linear behavior is solvable with only matrix multiplicatidns after
the first iteration and, consequently, problems of considerable
complexity can be finished on small computers readily available to
design engineers and geologists. Also, the load transfer method
allows computation of shear deformations after the peak load. As
discussed in Chapter 8, incremental loading and unloading is a good
procedure which will improve convergence in many cases.

* Professor Edward L. Wilson, "Analysis of Axisymmetric Solids",


University of California, Berkeley, Department of Civil Engineer-
ing, SESM Computer Programming Series (February, 1967) --
Subroutine BANSOL.
Appendix 1 371

The list of input parameters describing properties of discon-


tinuities has been kept short by coding much of the "model" in the
body of the program. These assumptions, discussed in Chapter 8, are
reviewed in Table A-I. Minor programming will be necessary to change
the details of the discontinuity model. Major programming may be
desirable to enlarge the scope in other ways. For example, a program
like JETTY can serve as the static elasticity "module" with other
subroutines to solve problems in coupled fluid flow, coupled heat
flow, dynamics, non-linear rock behavior, etc. One simple way to
enlarge the scope so as to permit study of construction sequence or
incremental loading type problems is to introduce modifications
manually when restarting continuing problems.
We now will consider the structure of the program in comparison
with the material presented in Chapter 8. To facilitate reading
the listing, equations are identified in numerous comment cards.

PROGRAM STRUCTURE
Table A-2 shows the sequence of computations and subroutine
calls for JETTY. The main program reads the input data and calls
subroutine STIFF to assemble the structural stiffness (KS) and
load vector {R}; this is done with the aid of subroutines TRIA and
JSTIF for triangles and joints respectively. The structural stiff-
ness matrix is inverted by a library matrix inversion routine (not
listed) which uses Gauss Elimination. The inverse of the stiffness
matrix, including the boundary conditions, is stored in the array
(KS). The main iterative loop of the program is entered after
initializing the displacements {u}. The total displacement at this
point becomes {u} = (KS)*{R}. From these displacements, solid
stresses are calculated in STRESS while joint deformations, and
initial load corrections, are computed in JSTR. Initial force
increments for the next iteration are returned by JSTR in {R},
and the iterative loop is re-entered with updated displacements
{u} = (KS)*{R} + {u}.
The sequence of computations and the equations programmed are
almost identical to those presented in Chapter 8 except for the
following points. Rather than compute an updated initial load vector
Good man-Geol og ical En g - 25
372 Append~ 1

to begin each new iteration, an incremental load vector is used and


the displacements computed from each iteration are added to those
previously computed. The incremental initial load vector will
approach zero if convergence is neared. Another point of difference
is the way in which the displacement constraints are introduced.
Rather than partition the stiffness matrix by removing the rows and
columns of known displacements, the stiffness term corresponding to
the row and column of a fixed nodal point displacement is set to the
20
arbitrarily high value of 10 . A non-zero displacement condition,
i.e. a moved and then fixed nodal point can be input by reading into
the load vector {R}, at the constrained node, the value of the given
20
displacement multiplied by L0 . Modified joint elements (figure
8-12) are not explicitly identified in the computation. Instead,
the left and right halves of each joint element are considered
separately in JSTR. Thus, each nodal point pair (a,b) of a modified
joint element receives its initial force increment computations first
from the half element on one side and then from the half element on
the other side. Limits have been set on k:
n 0.01 q
u IV mc -< k n-
<
100. qu/Vmc'
No test for convergence has been programmed. Problems with
diverging or unstable results can arise when stepping out very far
from an initial equilibrium, in which case loading or unloading will
have to be done in increments, with stress output from any increment
introduced as residual stress to begin the next increment. JETTY
has no provision to do this automatically. Numerical instability
can result when individual triangles are connected to more than one
joint, since incompatible load corrections may be required.
The program includes a number of WRITE statements intended for
debugging purposes. These will be executed if the word INTERMEDIATE
is introduced as an optional control card. Thus, a complete printout
of intermediate computations can be obtained whenever desired.
Samples of such output for examples 3 and 4 are presented later. The
intermediate results make the output long but will enable you to
follow the program logic.
Appendix 1 373

TABL E A-1

Summary of Assumed Material Property Relationships

ROCK

Linear, transversely isotropic solid. The rock elements may


represent uniformly bedded or slabby rock. There is no provision
for failure of the rock elements.

JOINTS

Asperities obey Fairhurst's failure criterion, consisting of


a parabolic envelope fitted to Mohr circles for the unconfined
compression and tension tests (Fairhurst, 1964).

Peak shear strength T is given by Ladanyi and Archambault's


p
equation with the transition pressure aT = qu of the wall rock
(equations 14 to 16 of Chapter 5).

Dilatancy v and area of contact as vary with normal pressure


as given by Ladanyi and Archambault, with aT quo

Residual shear strength Tr B(a)T p where B(a) decreases


linearly from Bo at a = ° to 1 at a = qu (60)*
The ratio of peak to residual displacements, M, measured in a
test beginning from T = 0, equals 4.
Normal displacement resulting from normal compression or
decompression obeys a hyperbolic law (43) or (49) with the initial
stiffness k determined by the initial stress (44a).
n
Dilatancy does not occur when the shear deformation exceeds
the residual value u . ( 61) an d (62).
r

* Numbers in parentheses refer to equations of Chapter 8.


374 Appendix 1

TABLE A-2
Structure of the Computer Program

MAIN PROGRAM (JETTY)


- Read data
Preprocess data; e.g. the global stress strain matrix is
assembled from the material properties.
- Assemble the stiffness matrix: CALL STIFF
SUBROUTINE STIFF
- For each element, compute the element stiffness matrix
and the initial loads. For solids, CALL TRIA; for joints,
CALL JSTIF
- Assemble the structural stiffness matrix and net initial
load vector {R}
- Introduce constraints
Return
- Solve for displacements: call a suitable equation solver to
invert the stiffness matrix. Return the inverse in (KS).
- Initialize displacements {u}.
- ENTER ITERATIVE LOOP
Compute incremental displacements and add to {u}: {u}
(KS){R} + {u}
- Compute solid stresses from known displacements: CALL STRESS
SUBROUTINE STRESS
For each solid element in turn:
- Assemble total nodal point displacements
- Form the element stress-displacement matrix
- Compute the element stress change and add to the element
initial stresses.
Return
- Compute joint stresses and define new initial load increments
{R} CALL JSTR
SUBROUTINE JSTR
For each joint element in turn:
- Assemble the total nodal point displacements
- Compute joint deformations in local coordinates
- Compare with the theoretical model for joint closing,
opening, shearing, and dilatancy.
- Compute corrective force increments and store for the
next iteration as equivalent initial stresses
- Transform force increments to global coordinates and
store in {R}.
Return
- Write output displacements and stresses
- Leave iterative loop if the last iteration
- Write restart information on tape
End
Appendix 1 375

INPUT INSTRUCTIONS FOR JETTY

The follow in g car ds constitute the data deck:

1. TITLE CARD
An y title message desired; it will h ea d the out p ut .
2. FIRST CONTROL CARD (7I5)
Field:
~umber of n o da l points
(2) number of elements
(3) numb e r of n o d al points in the list of fi xed po in ts
(4) number of solid material types
(5) number of joint material types
(6) number of the first iteration of this r un ( prev io us + 1 )
(7) number of iterations to be computed this run.
3. SECOND CONTROL CARD ( 5F IO.O)
Field :
~.O if gravity forces are to be computed ; ot he rw ise b lan k
( 2) t he accele r at ion of g r avity
(3) pseudo-stati c acceleration coefficient k (ac ce le r a tio n = kg )
(4) di rection of the inertia force of the applied ac celerat ion
a n angle in degrees counter- clockwise from x.
(5) the mass density of water (t h is establishes t he sys tem of
dimensions)
4. Data se t s foll ow i n any order. A hea d ing car d pre ce des e a ch
set and identifies it.
HEADING CARDS FOR DATA SETS; begin in column 1; 9-72 a r e optional.
NODAL POI NT DATA
ELEMENT DATA
HYDRAULIC DATA
RESIDUAL STRESS DATA
BOUNDARY CONSTRAI NTS
SOLID ELEMENT PROPERTIES
JOINT ELEMENT PROPERTIES
INTERMEDIATE PRINTOUT DESIRED (no data set f ollows thi s car d)
a) NODAL POINT DATA ( I 5,4FIO.O)
F i eld :
(1) nodal point number
(2 ) x coordinate
(3) y coordinate
( 4 ) applied ext ernal for ce in x d irection
(5) applied e x ternal force in y direction
Omitted nodal points a r e linearly in t e r pol a t ed wi t ho u t exte r n a l
forces. The last nodal point (highest numbe r e d ) mus t be inpu t.
b) ELEMENT DATA ( 61 5)
376 Appendix 1

Field:
(1) element number
(2) material number (joints are assigned numbers higher than
solids) .
(3) nodal point I
(4) nodal point J
(5) nodal point K
(6) nodal point L (blank for solid elements)
Nodal points are designated in counter-clockwise sequence. In
joints, the long sides must be IJ and KL (see figure 8-20).
Elements may be omitted; the numbers of omitted elements are
assigned by incrementing the corner numbers of the previous
element and the previous material number is assigned. The
highest numbered element must be input.
c) HYDRAULIC DATA (I5,3FIO.O)
Field:
(1) element number
(2) "Head" - Head is 0 for a solid element above the water table
and 1.0 for a solid element below the water table.
(3) ah/ax (negative for flow in the direction of positive x)
(4) ah/ay (negative for flow in the direction of positive y)
If elements are omitted from the list, the gradients will be
assigned equal to those on the preceding card. The highest
numbered element to receive hydraulic data must be input.
d) RESIDUAL STRESS DATA (I5,3FIO.0)
Field:
(1) element number
(2) (J
X,o
(3) (J y,o
(4) T
xy,o
Elements may be omitted, in which case the residual stresses will
be assigned equal to those on the preceding card. The highest
numbered element to receive residual stress must be input.
e) BOUNDARY CONSTRAINT DATA (8(215))
(1) nodal point
(2) 1CODE .....
if 1CODE = 0, no constraint is introduced (this is a
convenience when running multiple problems with changing
boundary conditions)
if ICODE 1, x displacement is zero ("rollers" parallel to
y)
if 1CODE 2, y displacement is zero ("rollers" parallel
to x)
if ICODE 3, x and y displacements are both zero (fixed)
(3) next nodal point
(4) 1CODE for next nodal point, etc. up to 8 nodal points per card.
At least one node must be constrained
AppendIx 1 377

f) SOLID ELEMENT PROPERTIES (I5,7FIO.O)


Field:
(1) material number; materials must be numbered in sequence
starting from 1.
(2) mass density
(3) Es - modulus of elasticity in direction parallel to s
(4) En - modulus of elasticity in direction parallel to n
(5) G - shear modulus in the sn plane
sn
(6) v Poisson's ratio giving strain in the n direction due
sn stress applied parallel to the s direction
(7) v - Poisson's ratio giving strain in the t direction due
st to stress applied parallel to the s direction.
(8) a --direction of the s axis, measured counter-clockwise
from the x axis (see figure 8-4)
g) JOINT ELEMENT PROPERTIES (I5,7FIO.O,F5.0)
Field:
(1) material number
(2) q the unconfined compressive strength of the asperities
u (negative)
(3) T /q - the ratio of tensile to compressive strength of wall
o u rock
3
(4) k - joint shear stiffness (dimensions force/length )
s
(5) B - the ratio of residual to peak strength at low normal
o
stress
(6) V - the maximum amount a joint can close from an initial
mc
seating load (positive)
(7) ~ - the seating load for measuring V (negative)
mc
(8) ¢~ - the friction angle for a smooth joint
(9) i the dilatancy angle at zero normal pressure
o
5. FINAL CONTROL CARDS (beginning in column 1)
First Card: (one of the following)
START - will execute program, forming the structural stiffness
matrix from the input information
STOP - will stop before executing program; no further card is
needed.
RESTART - will continue a previous problem without accelerating
the solution, reading required continuation data from tape 7.
In this case, the data cards can be omitted, i.e. the entire
deck consists of the title card (1) and the first and second
control cards (2) and (3), followed by RESTART and the last
control card. The structural stiffness matrix will not be
computed, but will be read from tape 7.
Last card (one of the following):
STOP - continuation information will not be written on tape at
the end of the problem.
SAVE STIFFNESS - continuation information will be written on
Append~ 1

tape at the end of the problem.


ACCELERATE - will write new stiffnesses, corresponding to a
variable stiffness iteration, on tape at the end of the
problem. On restarting the problem, the full data deck
must be used, and the first final control card must be START

Examples of restart information:


a) A run followed by a normal restart. (It will be necessary
to use the appropriate control cards to get and attach tape 7).
First run:
Standard data set
START
SAVE STIFFNESS
Second run:
Title card + first and second control cards
RESTART
STOP
b) A run followed by an accelerated restart (control cards
must be used to get and attach tape 7)
First run:
Standard data set
START
ACCELERATE
Second Run:
Standard data set
START
STOP

EXAMPLES OF INPUT AND OUTPUT

Example 3
Data coding form 1 shows the input information for example 3
of Chapter 8 (figure 8-10). In order to show the complete stiffness
matrix in the output, no nodal points were fixed; however since
at least one nodal point must be assigned a boundary constraint,
node 1 was assigned ICODE equal to zero (leaving it free). The data
deck includes the card INTERMEDIATE which causes all print statements
to be executed. Output number 1 shows the formation of element stiff-
ness matrices and the assembly of the structural stiffness matrix
and load vector.

Example 4
The input information for example 4 of Chapter 8 (figure 8-19)
is listed on coding form 2. Note the use of heading cards and the
intentional omission of some residual stress cards which will be
Append~ 1 379

generated inside the program. Computer output number 2 shows the


results of the first, second, fifth, and tenth iterations. The output
gives the total displacements and solid element stresses in global
coordinates (x,y) and joint deformations and stresses in local co-
ordinates (s,n). Remember that tension is positive; the minor
principal stress 03 will be the largest compression. The "orienta-
tion" given is the direction of 01 measured counter-clockwise from
x. Output number 3 is a complete printout with intermediate results
for computation of the first iteration.
~ ~
t-3 C
:z>
()
o
t:I
H

~
I-xj

@
a=
I-l

,
I ,

b_U ~ h I" Cblrd 1; I : ' 1 ' 1 ' : : I : : : 1: I : ! i [ 1I : : ! I


!:tIN T£ QI~£Dz~trE
INlfl)~~ Ip0~NIT DAil..
II I 01. 0,.

'I I
21 I
31 I
~II
SI I
1,.
£ILE""EIMtr .D~tr~
()I.
0,.
2
2,.
0,---
, I. '
'0
0,. :E ,
I
: I

I
'
I

~ l' h'
i
II I 11+1 , 5 i I I

i I
I I •

I
21 I '1 21 '51 "3 J I

~L Z ~ I 'I ' 'I 3~ t


I
:1
II I
E L£ME NT
01 •
P R<lPE~
, 10 0
; :5
F-Sa
() •
t.;,;
1 CICIO 41' 7. 01.2. 01·1 o.
JI"IN Tl IELENIEAlT pjR"PERrr'!.ES -
21 I -IiCoo. I +01. , '5100.' ~., ,01.005 - '1. :0 'i:o:. ! 1
!
51.0
8IoU 'Nl>IAIRY G~HI>'[N;N~
j
!
I I
II I 0 ! '
JlIESJ:.~"~ L
I
SI"TR!'551
-'0.0
D"iAt------=-
-51.0 - • 1.0 )::,.
:g
/1 31 / - '10' 0 / - 5/. 0 - 11.0
I __
~
~T~~TI t:t
IstTo'P )('
I I '1 .5 16 I; e 9 to II 11 I'J I.: 15 16 17 18 19 20 71 n 13 1·: 2~ 26 n 26 29 30 3 1 32 33 34 35 )6 3738 39 dO .41 .;2 ..:) ':4 .as 46 47 .as .9 50 51 .52 5J 5.010 55 56 ;7 58 ;9 60 61 67 63 64 65 66 67 68 69 70 71 n 173 7.: 75 76 n 78 79 BO
t:::l :t:.
:J>
1-3
:g
CtI
:J> :::3
(j
t:t.
)<'
0
t:::l
H
~
Q
f'%j
0
::d
;s::
~
I 23 .i S 6 - 8 Q HI 1., 11 .~, H ''II 1:'1 ~~ ?J 7: ] J 26; lB i o; 10 l' 31 2.1 J;: 1', 11'1 I, 1'4 " ,I I 1. 1 ,~ / .!I .,;.: \', ~~.:Io <1 950 5 1 05' 5 J ~J. 55 56 57 s.a :59 60 ci to ~ ~3 6': M be t. 7 '..8 69 711 71 i' ';' n n. 15 7677 ia 79 en
j ~~p ~~~ J"X,.,,r c. L. siW6-
, IH I)~ 1t-,":M'Jit R £ ' 5 'S~ ,1J IN 0..- ' IA..M r: w~ T IA. L L. Y ST ICl,1a. 5 S :£ [I) . ' Ia,L GaC :tc
I I

7 S I I 5
b Jahk c:::._~_~ IA
"'4b~L
, P ~ r.hl IT
0,
.b~'T
"'- 0,
-- I --

I-
-- -

Z , .0 O .
~ I .0 0. '
~ 0. 5 0, '5 -
S' o. , .0
--
6 l ,. O I .0
-
7 '. 0 I .0
ELf"'E NT
I ,
D A T A. -
I
--
1 ...
-- - -
: '

:1
3 ,
J
,
2. 6
5
4-
4-
~

,
I ' I '

t- I - i- _ I - 5 J ... -
5 .2. 3 7 6 2 -
Re.Sll)
, UA~
-5
~TRe. SS
.0
b A.,. A.
I
,
~ - t .0 -
a,UNO AiR,y
, c c6,. D J:T le NS
3 3 ; ~ S 5 3 6 2- 7 ~_
I
I

S 95LI. 1) EL£ MEtJr PRt5PE. ~ tT':L--'u_ -


I '0 0 . I~O. 5b,
Ji4I.NT E.I-5.H EN T PI _ ~ p E.~ rr.t.E.s, I

~ - J 00. +_C) . I - - 1 0 0. Cl.6 0.0; - 0 . I 30 .


START
~T0C)
-
I 2 J ..\ 5 6 i 8 9 10 11 I ~- 11 1£ IS r. . 17 Il:I 19 J(i ~ /} lJ '~ ~ __ ?!_1; IH Iq lIJ JI j7 J;i J~ L ]0 )7 )8 j9 .to .:.1 " 1 4' .tJ. .is .016 J.;- ;:,8 "'9 !-O ~I ~,' ,J -...l ', ', I to '1." __ ,II ;"Q 60 e.I t-1 6J 6.J ,,5 ~ 67 b8 6 Q :0 " n 7J 7·1 /516 7/1 19 7980
~
EXAMPLE NUMBER 3 WITH COMPLETE PRINTOUT OF INTERMEDIATE RESULTS

NUNBER OF NODAL POINTS 6


NUMBER OF ELEMENTS :3
NUNBER OF CONSTRAINED NODES=
NUMBER OF SOLID MATERIALS
NUMBER OF JOINT TYPES
NUMBER OF FIRST ITERATION I
NUMBER OF ITERATIONS 1
GRAVITY LOADING - ON/OFF O.
ACCELERATION DUE TO GRAVITY= O~
PSEUDO STATIC ACCELERATION O.
DIRECTION OF ACCELERATION -0.
MASS DENSITY OF WATER -0.

NODE XORD YORD XLOAD YLOAD NODE XORD YORD XLOAD YLOAD
1 o. O. O. O. 2 O. O. O. O.
3 O. 1.00 O. O. 4 2.00 O. O. O.
5 2.00 O. O. O. 6 2.00 1.00 O. O.

ELEMENT MAT TYPE NODE LIST ELEMENT MAT TYPE NODE LIST
1 2 I 4 5 2 2 1 2 5 3 0
3 6 3 5 0

SOLID MATERIAL PROPERTIES


MATERIAL TYPE
MASS DENSITY O.
YOUNGS MODULUS - PARALLEL TO ANISOTROPYa .1000E+04
YOUNGS MODULUS - NORMAL TO ANISOTROPY .1000E+04
SHEAR MODULUS - GSN .4170E+03
POISSONS RATIO - SN .2000
POISSONS RATIO - ST .2000
ORIENTATION OF LAYERS (ANGLE S-X) O.

STRESS-STRAIN MATRIX IN LOCAL CO-ORDINATES


.IIIIIE+04 .27776E+03 O•
• 27778E+OJ .11111E+04 O.
O. O. .41700E+03

TRANSFORMATION MATRIX
• 10000E+Ol O. O.
O. .10000E+Ol -0.
O. -0. .10000E+OI

STRESS-STRAIN MATRIX IN GLOBAL COORDINATES


.11111E+04 .27778E+03 O•
• 27778E+OJ .11111E+04 O.
O. O. .41700E+03

JOINT PROPERTIES
JOINT TYPE NUMBER 2
WALL ROCK COMPRESSIVE STRENGTH -.1000E+04
RATIO OF TENSILE TO COMPRESSIVE STRENGTH= .1000
SHEAR STIFFNESS .5000E+OJ
RATIO OF RESIDUAL TO PEAK SHEAR STRENGTH: .6000E+00
MAX I MUM NORMAL CLOSURF .5000E-C2
SF:ATI"'G LOAD -.1000E+Ol
FqlCTION ANGL[ OF A S~OOTH JOINT .3000E+02
DILATANCY ANGL~ .500JE+Ol
ELEMENT QF Sl ou~L S TRESSES XX, YY, XY ELEME NT RESIDUAL STRESSE S XX,YY,X Y
J -.IO OOE+ 0 2 -.S OOOE +OI -.I OO OE+Ot 2 -. 1000E+02 -.SO OOE +OI - . IOOOE+OJ
3 -.IOOOE + 0 2 -.SOOOE+OI -.IO OOE +OJ

NODE D IS PLACE~E NT CODE


I - 0

JOINT RE SIDUAL SHEAR STRESS - . IOOOE+O I


JO t NT RE SIOUA L NORMAL STRESS= - . 5000E +Ot

J OI NT S TI FF NE S S - LOC AL COORD I NA TE S
.2500 E+03 O. . 250 0 E +0 3 O. -.2500E+03 O. -.2500E+03 O.
O. • 5000f +O'" O. O• O. O. O. -.5000E+0~
.2500E "0 3 O. • 2500E+03 O. -.25 00 E +0 3 O• -.250 0E .. 03 O.
D. O. O. • 5000E+04 O. - .S OOO E +04 O. O•
-.2500E +03 O. - . 2500E"03 O. . 2500E+ O.J O. . 25001:+03 O.
O. O. O. -. 5 00 0E+ 04 O. . S OO OE +O", O. O.
-.250CF "03 O. -.2500E+03 O. . 2500E+ 0 3 O. . 2500E+'J3 O.
o. -.5000E+04 O. O. O. O. O. . 5 000E+ 0~

JOINT S T IF F NES S AND FORCE


. 25 0 0E +-03 O.
VECTOR - GL OBAL COORDS.
• 2S 0 0 E+0 3 O.
-
ELE14E NT
- . 2 50 0 E+ 03
I
O. -. 2500E+03 O• -. IOC OE+OI
O. • 50 00E+04 O. C• O. O. O. - . 5000E+04 -.5000E+OI
. 2500E +03 O. . 2500E+03 O. -.2500E+03 O. -.2500E+03 O. -. IO OOE+O I
O. O. O. . 5000E +04 O. -.5000E+04 O. O. -. 5000E+O I
-.2500 E +03 O. -. 2 5 00 E+03 O. .2500E+ 03 O. . 2500 E+0 3 o. . 1 000E+Ol
O. O. O. -. 5000E+04 O. .5000E+04 O. O. . 5 0 00E +O I
- .2500E+03 O. - .2500 E +0 3 O. .2500E~ 03 O. . 2 500E +0 3 O. .1 000E +O l
O. -.5000E+04 O. O. O. O. O. .5000E+04 .50 OOE+O I

ST RAI N DI S P L ACE ME NT MATRIX


- . 5 0 00E+00 O. .50 00E+00 O. -0. O.
O. - .IOaOE +OI O. -0. O. .1000E+Ol
-.1000E+Ol -.5000E+ 00 - 0 . .5000E+00 0100 0E +OI -0.

STRESS-DISPLACEMENT TRANSF ORMATI ON MAT R IX


-.5556E+03 -. 27 78£+ 03 . 5556 E+0 3 O. O. .2778E+ 03
- .1 389E" 03 - . 1 111 E +04 .i389E+0 3 O. O. .111IE +0.
-.4170E+03 - .2085E + 03 o. . 2085E+03 .~17 0E+ 03 O.

ST IF F NESS MAT RIX


.6948 E +03
-EL E~ E NT
.34 7 4E+03
2
- .2 778E+03 - . 2 08 5E+ 03 - .~1 70E +03 -. 1 389E+ 03
L OAL) VE CTO R
-.600 0 E +OI
. 34 7~ E +03 . 1 215E+ 04 -. 1389E*0 3 - .1 0 4 2 E+0) - .2085E +0.3 -.IIII E +04 -. 550 0 E +OI
- .277 8E+ 03 -.1389E "03 • 27 78F "0 3 O• O. .13 89E+03 .500 0E+ Ol
-. 2 0 8 5 E +03 - . 104 2E+0 3 O. . 104 2 E +03 • 20 8 5E+03 o• . 5 000 E-.-0 0
- .417 0 E .03 - . 2 0 8 5E+0 3 O. • 20 85E +03 •• ,70E+O) O• • 100O E+Ol
-.1389 E +03 -.1I11E+04 .1389E+0 3 O. O. . IIIIE+ 04 .500 0E+OI

ST RAIN DISPLACE ME NT MAT R I X


.SOOOE +OO O. - . 5 000E +0 0 O. -0. O.
O. . 100 0E+Ol O. - 0. C. - .100 0E+ Ol
• I OOOE + OI • 500 0 E +00 -0 • -.50 00E+0 0 - .1000E+Ol -0 •

STRESS- DI S PLAC ENENT TR ANS F OR MATI ON ~A T Rt X


. 5556E +03 .2 778E +03 - .555 6E+ 0 3 O. O. - . 2 7 78E+03
• 1 3 89E+0 3 . 111 1E+04 -. 1 389E+ 03 O• O. -.1111E+04
.4170E+03 • 208 5E +03 O. -.2 0 85E+03 -.41 70E +03 O•
-
~
STIFFNESS MATRI)( ELE'-'ENT 3 LOAD VECTOR
.694I3E+03 .3474E+03 -.2778E+03 -.2085E+03 -.4170E+03 -.1389E+03 .6000E+Ol
.3474E+03 .1215E+04 -.1389E+03 -.1042E+03 -.2085E+03 -.IIIIE+04 .5500E+Ol
-.2778E+03 -.1389E+03 .2778E+03 O. O. .1389E+03 -.5000E+Ol
-.2085E+03 -.1042E+03 O. .1042E+03 .2065E+03 o. -.5000E+00
-.4170E+03 -.2085E+03 O. .2085E+03 .4170E+03 O. -.1000E+Ol
-.1389E+03 -.IIIIE+04 .1389E+03 O. o. .1111E+04 -.5000E+Ol

LOAD VECTOR AND STIFFNESS MATRIX


• 2500E+03 O. -.2500E+03 O. O• O. • 2500E+03 o. -.2500E+03 O•
O. O.
O. • 5000E+04 O. -.5000E+04 O. O• O. O. O. O.
O. O.
-.2500E+03 O. .9448E+03 .3474E+03 -.4170£+03 -.1389E+03 -.2500E+03 O. -.2778E+02 -.2085E+03
O. o.
O. -.5000E+04 .3474E+03 .6215E+04 -.2085E+03 -.1111E+04 O. O. -.1389E+03 -.1042E+03
O. O.
O. O. -.4170E+03 -.2085E+03 .6946E+03 O. O. O. O. .3474E+03
-.2778E+03 -.1389E+03
O. O. -.1389E+03 -.1111E+04 O. • 1215E+04 O• O. .3474(+03 O.
-.2085E+03 -.1042E+03
• 2500E+03 O. -.2500E+03 O• O. O. .2500E+03 O. -.2500£+03 O.
O. O.
O. O. O. O. O. O. O. .50'OE+04 o. -.5000E+04
O. O.
-.2500E+03 O. -.2778E+02 -.1389E+03 O. • 3474£+03 -.2500E+03 O• • 9448E+03 O•
-.4170E+03 -.2085E+03
O. O. -.2085E+03 -.1042E+03 • 3474E+03 O• O. -.5000E+04 O. .6215E+04
-.1389E+03 -.1I11E+04
O. O• O. O. -.2778E+03 -.2085E+03 O. O. -.4170E+03 -.1389E+03
.6948E+03 • 3474E+03
O. O• O. O. -.1389E+03 -.1042E+03 O. O. -.2085£+03 -.1111E+04
.3474E+03 • 1215£+04
-.1000E+Ol -.5000E+Ol -.5000E+Ol -.5000e+Oo -.4000£+01 .4500E+Ol -.1000E+Ol -.5000e+Ol .5000£+01 .5000E+00
.6000E+Ol .5500e+Ol
EXAMPLE 4. JOINT CLOSING BY DECOMPRESSION OF AN INITIALLY STRESSED ALOCK

NU~9ER OF NODAL POINTS 7


NUMBER OF ELEMENTS 5
NUMBER OF CONSTRAINED NUDES= 6
NUMBER OF SOLID ~ATERIAL5 I
NUMBER OF JOINT TYPES
NU~BER OF FIRST ITERATION
NUMBER OF ITERATIONS 10
GRAVITY LOADING - ON/OFF -0.
ACCELERATION DUE TO GRAVITY= -0.
PSEUDO STATIC ACCELERATION -0.
DIRECTION OF ACCELERATION -0.
MASS DENSITY OF WATER -0.

NDDE XORD YaRD XLOAD YLOAD NOf)E XORD yaRD XLOAI.> YLOAD
I -0. -0. -0. -0. 2 1.')0 -0. -0. -0.
3 1.00 -0. -0. -0. 4 • 50 • 50 -0 • -0 •
5 -0. 1.00 -0. -0. 6 1.00 1.00 -0. -0.
7 1.00 1.00 -0. -0.

ELE"IENT MAT TYPE NODE LIST ELE"4ENT "4AT TYPE NODE LIST
I I 2 4- -0 2 I 2 6 4 -0
3 6 5 4- -0 4- 5 4 -0
5 2 3 7 6 2

ELEMENT RESIDUAL STRESSES XX,YY,XY ELEMENT RE 51 OUAL STRESSES XX,YY,XY


I -.5000E"01 -0. -0. 2 -.5000E+-OI -0. -0.
3 -.5000E"01 -0. -0. 4. -.5000E+-OI -0. -0.
5 -.IOOOE+-Ol -0. -0.

NODE DISPLACEMENT CODE


I 3
2 2
3 3
5 3
(, 2
7 3

SOLID MATERIAL PROPERTIES


MATERIAL TYPE
MASS DENSITY -0.
YOUNGS "IOOULUS - PARALLEL TO ANISOTROPY= .1000E+-03
YOUNGS MODULUS - NORMAL TO ANISOTROPY .1000E+-03
SHEAR MODULUS - GSN .5000E+-02
POISSON5 RATIO - SN -0.
POISSONS RATIO - ST -0.
ORIENTATION OF LAYERS (ANGLE S-Xl -0.

JOINT PROPERTIES
JOINT TYPE NUMBER 2
WALL ROCK COMPRESSIVE STRENGTH -.1000E+-03
RATIO OF TENSILE TO COMPRESSIVE STRENGTH= .1000
SHEAR STIFFNESS .1000E+-03
RATIO OF RESIDUAL TO PEAK SHEAR STRENGTH= .6000E+00
MAX I MUM NORMAL CLOSURE • 5000E-0 I
SEATING LOAD -.IOOOE+-OO
FRICTION ANGLE OF A SMOOTH JOINT .300I)E+-02
DILATANCY ANGLE -0.
ITERATION NUMBER

NODE X-DI SPLACEMENT Y-DISPLACEMENT


1 -.000000 .000000
2 .013333 -.000000
3 .000000 -.000000
4 .006667 .000000
5 -.000000 .000000
6 .013333 -.000000
7 .000000 -.000000

SOLID ELEMENT STRESSES

ELEMENT XX-STRESS YY-STRESS XV-STRESS I<4AJOR PRINCIPAL MINOR PRINCIPAL ORIENTATION


1 -.3667E+Ol .9715E-14 .2842E-13 .7105E-14 -.3667£+01 90.16
2 -.36671'0+01 .4815E-32 .2356E-13 O. -.3667E+Ol QO.16
3 -.3667E+Ol -.9715E-14 O. -.7105E-14 -.3667E+Ol 90.16
4 -.3667E+Ol -.6019E-32 .4857E-14 O. -. 3667E+0 1 90.16

JOINT DISPLACEMENTS JOINT STRESS~S


ELEMENT SHEAR NORMAL ROTATION SHEAR NORMAL MOMENT
5 -.2407E-34 -.1333E-Ol -.6106E-15 -.2407E-32 -.3667E+Ol -.3053E-13

ITERATIQN NUMBER 2

NODE X-DI SPLACEMENT Y-D 1 SPLACEMENT


1 -.000000 .000000
2 .006869 .000000
3 .000000 .000000
4 .003434 .000000
5 -.000000 -.000000
6 .006869 .000000
7 .000000 .000000

SOLID ELEMENT STRESSES

ELEMENT XX-STRESS YY-STRESS XV-STRESS MAJOR PRINCIPAL MI/ltOR PRINCIPAL ORIENTATION


1 -.4313E+Ol 01021E-18 .8882E-14 .1421E-13 -.4313E+Ol 90.16
2 -.4313E+Ol .2407E-32 .5329E-14 O. -.4313£+01 90.16
3 -.4:113E+Ol -.1021E-18 -.3553E-14 -.1421E-13 -.4313E+Ol -90.16
4 -.4313E+Ol -.3009E-32 .5104E-19 O. -.4313E+01 90.16

JOINT DISPLACEMENTS ..101 NT STRESSES


ELEMENT SHEAR NORMAL ROfATION SHEAR NORMAL MOMENT
5 .3009E-34 -.6869E-02 -.8327E-16 .3009E-32 -.4313E+Ol -.3081E-13
ITERATION NUMBER 5

G">
g
~ NODE X-DISPLACEMENT Y-OI SPLACEMENT
[ I
2
-.000000
.00~009
-.000000
.000000
_.
3 .000000 .000000
'"en .002005 -.000000
"5 -.000000 -.000000

~ 6
7
.004009
.000000
.000000
.000000

SOLID ELE~ENT STRESSES

ELEIoIENT XX-STRESS VV-STRESS XV-STRESS "'''JOR PRINCIPAL MINOR PRINCIPAL ORIENTATlON


I -.4599E+OI -.158lE-14 .l55JE-14 -.142IE-13 -.4599E+OI 90.16
2 -.4599E+OI .1505E-32 -.9671E-16 o. -.~599E+OI -90.16
3 -.4599E+Ol • 1583E- 14 -.444IE-14 .I~2IE-13 -.4599E+OI -90016
4 -.4599E+OI -.1749E-32 -.7915E-15 O. -.4599E+OI -90.16

JOINT DISPLACEMENTS JOINT STRESSES


ELEMENT SHEAR NORMAL ROTATION SHEAR NORMAL MOMENT
5 • J385E-34 -.4009E-02 .2776E-16 .3385E-32 -.4599E+OI -.2365E-13

ITERATION NUIoISER 10

NODE X-OISPLACE~ENT Y-DISPLACEMENT


I -.000000 -.000000
2 .003915 .000000
3 .000000 .000000
4 .001958 -.000000
5 -.000000 -.000000
6 .003915 .000000
.000000 .000000

SOL 10 ELE~ENT STRESSES

ELE~ENT xx- STRESS YY-STRESS XV-STRESS MAJOR PRINCIPAL MINOR PRINCIPAL ORIENTATION
-.4608E+OI -.7697E-15 .8882E-15 -.142IE-I] -.4608E+OI 90.16
2 -.46 08E+0 I .147IE-32 • 1273E-I 4 O• -.~608Ef-01 90.16
3 -.4608E+OI • 7697E-15 O• .I42IE-I] -.~608E+OI 90.16
~ -.4608Ef-01 -.1 703E- 32 -.3849E-15 O. -.~608E+OI -90.16

JOINT DISPLACEMENTS JOINT STRESSES


ELEMENT SHEAR NORIoIAL ROTATION SHEAR NORMAL MOMENT
5 .2749E-34 -.3915E-02 -.2776E-16 .2749E-32 -.4608E+OI -.2226E-13
EXAMPLE -4 .JOINT CLOSING COMPLETE PRINTOU-T OF FIRST ITERATION

NU~BER OF NODAL POINTS 7


NUMBER OF ELEMENTS 5
NUMBER OF CONSTRAINED NODES= 6
NUMBER OF SOLID ~ATERIALS
~UMBER OF .JOINT TYPES
NUNBER OF FIRST ITERATION
NUMBER OF ITERATIONS 1
GRAVITy LOADING - ON/OFF -0.
ACCELERATION DUE TO GRAVITY= -0.
PSEUDO STATIC ACCELERATION -0.
OIRECTION OF ACCELERATION -~.
MASS DENSITY OF WATER -0.

NODE XORD YORD XLOAO YLOAO NODE XORD YORD XLOAD YLOAD
1 -0. -0. -0. -0. 2 1.00 -0. -0. -0.
1.00 -0. -0. -0. .50 .50 -0. -0.
:J
s -0. 1.00 -0. -0. "
6 1.30 1.00 -0. -0.
7 1.00 1.00 -0. -0.

ELEMENT MAT TyPE NODE LIST ELEIoIIENT MAT TyPE NODE LIST
1 1 2 -0 2 I 2 4 -0
J 1 6 5 " -0 5
6
4 -0
5 2 3 7 "
6 2 "
ELEMENT RESIDUAL STRESSES XX.YY.xy ELEMENT RES -I DUAL STRESSES Xl(.YYtl(Y
1 -.5000E+Ol -0. -0. 2 - .5000-E +0 I -0. -0.
3 -.5000E+Ol -0. -0. 4 - .5000EHl1 -0. -0.
5 -.1000E+01 --0. -0.

NODE DISPLACEMENT CODE


1 J
2 2
J :3
S 3
6 2
7 3

SOLID MATERIAL PROPERTIES


MATERIAL TYPE I
MASS DENSITY -0.
YOUNGS MODULUS - PARALLEL TO ANISOTROPY= .1000E+OJ
YOUNGS 1oII0DULUS - NORMAL TO ANISOTROPY .IOOOE+OJ
SHEAR MODULUS - GSN .5000E+02
POISSONS RATIO - SN z -0.
POISSONS RATIO - ST -0.
ORIENTATION OF LAYERS (ANGLE S-l() -0.

STRESS-STRAIN MATRIl( IN LOCAL CO-ORDINATES


.10000E+03 O. O.
o• • 1000 OE + 0 J 0•
O. O. .50000E+02

TRANSFORMATION MATRIX
.10000E+Ol O. O.
O. .10000E+Ol -0.
O. -0. .10000F+Ol
STRESS-STRAIN MATRIX IN GLOBAL COORDINATES
.lOOOOE+03 O. O.
O. ·lOOOOE+03 O.
O. O. .SOOOOE+02

JOINT PROPERTIES
JOINT TYPE NUMBER 2
WALL ROCK COMPRESSIVE STRENGTH -.lOOOE+03
RATIO OF TENSILE TO COMPRESSIVE STRENGTH= .1000
SHEAR STIFFNESS .1000E+03
RATIO OF RESIDUAL TO PEAK SHEAR STRENGTH= .6DOOE+OO
MAX I MUM NORIo4AL CLOSURE • SOOOE-O 1
SEATING LOAD -.IOOOE+OO
FRICTION ANGLE OF A SMOOTH JOINT .3000E+02
DILATANCY ANGLE -0.

STRAIN DISPLACEMENT MATRIX


-.lOOOE+Dl O. .1000E+Ol O. -0. O.
O. -.lOOOE+Ol O. -.IOODE+Ol O. .2000E+Ol
-.lOOOE+Ol -.lOOOE+Ol -.lOOOE+Ol .lOOOE+Ol .2000E+Ol -0.

STRESS-DISPLACE~ENT TRANSFORMATION MATRIX


-.lOOOE+03 O. .1000E+03 O. O. O.
O. -.lOOOE+OJ O. -.lOOOE+03 O. .2000E+03
-.SOOOE+02 -.SOOOE+02 -.5000E+02 .SCOOE+02 .lOOOE+03 O.

STIFFNESS MATRIX - ELEIo4ENT 1 LOAD VECTOR


.3750E+02 .12S0E+02 -.l2S0E+02 -.12S0E+02 -.2500E+02 O. -.l2S0E+Ol
.1250E+02 .3750F+02 .1250E+02 .12S0E+02 -.2S00E+02 -.SOOOE+02 O.
-.12S0E+02 .1250E+02 .37S0E+02 -.l2S0E+02 -.2500E+02 O. .1250E+Ol
-.1250E+02 .1250E+02 -.l2S0E+02 .3150E+02 .2500E+02 -.5000E+02 O.
-.2500E+02 -.2500E+02 -.2500E+02 .2S00E+02 .5000E+02 O. O.
O. -.SOOOE+02 O. -.5()OOE+02 O. .lOOOE+03 o.
STRAIN DISPLACEMENT MATRIX
.IOOOE+Ol O. .lOOOE+Ol O. -.2000E+Ol O.
O. -.lOCOE+Ol O. .1000E+Cl O. O.
-.IOOOE+Ol .1000E+Ol • lOOOE+Ol .1000E+Ol O• -.2000E+Ol

STRESS-DISPLACEMENT TRANSFORMATION MATRIX


.lOOOE+03 O. .lOOOE+03 O. -.2000E+03 O.
O. -.lOOOE+03 O. .lOOOE+03 O. O.
-.5000E~02 .SOOOE+02 .5000E+02 .SOOOE+02 O. -.lOOOE+03

STIFFNESS MATRIX - ELEMENT 2 LOAD VECTOR


.3750E+02 -.125CE+02 .l250E+02 -.1250E+02 -.SOOOE+02 .2500E+02 • 12S0E+O 1
-.l250E+02 .3150E+02 • 1250E+02 -.1250E+02 O. -.2S00E+02 o•
.1250E+02 .12S0E+02 .31S0E+02 .l250Et-02 -.5000E+02 -.2S00E+02 .1250E+Ol
-.l250E+02 -.1250E+02 • 1250E+02 • 3150E+02 O• -.2500E+02 O•
-.5000E+02 O. -.5000E+02 O. .100OE+03 O• -.2S00E+Ol
.2500E+02 -.2500E+02 -.2500E+02 -.2500E+02 O. • 5000E+02 O.

STRAIN DISPLACE~ENT MATRIX


.1000E+Ol O. -.lOOOE+Ol O. -0. o.
O. .1000E+Ol O. .100OE+Ol O. -.2000E+Ol
.lOOOE+Ol .1 OOOE+O 1 • 10OOE+Ol -.lOOOE+Ol -.2000E+Ol -0 •
STRESS-DISPLACEMENT TRANSFORMATION MATRIX
• 1000E.03 O. -.1000E+03 O. O. 0 .•
O. .1000E.03 O. .1000E+03 O. -.2000E+03
• 5000E+02 .5000f+02 .5000E+02 -.5000E+02 -.1000E+03 O•

ST IFFNESS MATRIX - ELEMENT 3 LOAD VECTOR


.3750E+02 .1250E+02 -.1250E+02 -.1250E+02 -.2500E+02 O. .1250E+OI
.1250E+02 • 3750E+02 .1250E+02 .1250E+02 -.2500E+0.2 -.5000E+0.2 O•
-.1250E+02 .1250E+02 • 3750E+02 -.1250E+02 -.2500E+02 O• -.1250E+Ol
-.1250E.+02 .1250E+02 - .1250E+02 .3750E+02 .2500E+02 -.5000E+02 O.
-.2500E+02 -.2500E+02 -.2500E+02 • .2500E+0.2 .5000E+02 O. O.
O. -.5000E"+02 O. -.5000E+02 O. .1000E+03 O.

STRAIN DISPLACEMENT MATRIX


-.IOOOE+OI O. -.IOOOE+OI O. .2000E+OI O.
O. .IOOOE+OI O. -.1000E+OI O. O•
• 1000E+Ol -.IOOOE+OI -.IOOOE+OI -.1000E+OI O. .2000E+OI

STRESS-DISPLACEMENT TRANSFORMATION MATRIX


-.1000E+03 O. -.1000E+03 O. .2000E+03 O.
O. .1000E+03 C. -.1000E+03 O. O•
• 5000E+02 -.5000E+02 -.5000E+02 -.5000E+02 O. .1000E+03

STIFFNESS MATRIX - ELEMENT 4 LOAD VECTOR


.3750E+02 -.1250E+02 .1250E+02 -.1250E+02 -.5000E+0? .2500E+02 -.1250E+Ol
-.1250E+02 .3750E+02 .1250E+02 -.1250E+02 O. -.2500E+02 O•
.1250E+02 .1250E+02 • 3750E+02 .1250£+02 -.5000E+02 -.2500E+02 -.1250E+Ol
-.1250E+02 -.1250E+02 .1250E+02 .]750E+02 O. -.2500E+02 O.
-.5000E+02 O. -.5000E+02 o. .1000E+03 O. .2500E+OI
.2500E+02 -.2500E+02 -.2500E+02 -.2500E+02 o. .'5000E+02 O.

JOINT RESIDUAL SHEAR STRESS = O.


JOINT RESIDUAL NORMAL STRESS= -.IOOOE+Ol

JOINT STIFFNESS - LOCAL COORDINATES


.2500E+02 O. .2500E+02 O. -.2500E+02 O. -.2500E+02 O.
O. • 1000E+03 O. O. O. O• O. -.1000E+03
.2500E+02 O. .2500E+02 O. -.2500E+02 O. -.2500E+02 O•
O. O. O. • 1000E+03 O. -.IOOOE+OJ O. O.
- .2500E+02 O. -.2500E+02 O. .?500E+02 O. .2500E+02 O.
O. O. O. -.1000E+03 O. .1000E+03 O. O.
-.2500E+02 O. -.250CE+02 O. .2500E+02 O. .2500 E+02 O.
O. -. 1000E+OJ o. o. o. O. O. .1000E+03

JOINT STIFFNESS AND FORCE VECTOR - GLOBAL COORDS. -O. ELEMENT 5


• 1000E+03 O. O. O. O• -.1000E+03 O. .5000E+00
O. • 2500E+02 O• .2500E.02 O. -.2500E+02 O. -.2500E+02 -0.
O. O. • 1000E+03 O• -.1000E+03 O. O. O. .5000E+00
O. .2500E+02 O. .2500E+0.2 O. -.2500E+02 O. -.2500E+02 -0.
O. O. -.1000E+03 O. .1000E+03 O. O. O. -.5000E+00
O. -.2500E+02 O. -.2500E+02 O. .2500E+02 O. .2500E+02 O.
-.1000E+03 O. O. O. O. O. .1000E+03 O. -.5000E+00
O. -.2500F+02 O. -.2500E+02 O. .2500E+02 O. .2500E+02 O.
h
LOAD VECTOR AND ST IFFNESS
.1000E+21 .2500E+02
MATRIX
-.1250E+02 -.12501:+0.2 O. O. -.7500E+02 -.2500E+02 .1250E+02 .1250E+02
:g
Cb
O. O. O. O. ~
.2500E+02 .1000E+21 .1250E+02 • 1250E+02 O• O. -.2500E+02 -.7500E+02 -.1250E+02 -.1250E+02 Q.
O. o. O. O. )<'
-.1250E+02 • 1250E+02 .1750E+03 -.2500E+02 -.1000E+03 O. -.7500E+02 .2500E+02 O• O.
· 1 2 50E+02 -.12 5 0 E + 02 O. O.
-.1 2 5 0 E +02 .12 50E + 02 - . 2500 E +0 2 • 1 000E +21 O. -.25 00 E + 02 O• O.
. 1250E+02 .1 250E + 0 2 O. - .2 50 0E+02
O. o. - . 100 0E+03 C. .1 0 0 0 E +21 O. O. o. O. O.
o. o. O.

-
O.
o. O. o. -.2500E +02 O. . 1000 E+21 O. O. O. o.
o. -. 2 500E+02 O. .25 00 E +02
- .7 5 00E +02 -. 2 500 E +02 -.7500 E + 02 • 25 0 0E+02 O. O. .3000E "03 O• -.75 00 E"02 .250 0E +02
- .7 500E + 02 -.2500E+02 O. O.
-.2500E+02 - . 7500E+02 .2 50 0 E+0 2 -.7500E +02 O. O. O. .3000E+03 .2 500E+02 -.7500E+02
-.2 500E+02 - . 7 500E+02 O. O.
• 1250E+02 -.1250E+02 O. O. O• O. -.7500E+02 .2500E+02 .1000E+ 2 1 -.2500E+02
-.1 2 50E+02 .1250E+02 O. O•
• 1250E+02 -.1250E +0 2 O. O. o. O. .2500E+02 -.7500E"02 -.2500E+ 0 2 .1000E"21
-.1250E+02 . 1250E+02 O. O.
O. O. . 1 250 E +02 .1250E +02 O. O. -.7500F+02 -.2500E+02 - .1250 E+02 -.1250E+02
• 1 750E+03 .2500E+02 - .1 0 00E+03 O•
O. o. -.125CE+02 .1250E + 0 2 O. -.2500E+02 -.25 0 0 E+02 -.7500E+02 .1250E"02 .1250E+02
.25COE+02 .1000E+21 O. -.2500E+02
O. c. O. O. O. O. O. O. O. o.
-.1000E+03 O. .1000E+21 O.
O. O. O. -.2500E+02 O. .2500E+02 O. O. O. O•
O. -.2500E+02 O. • 1000E+21
-.2500E+Ol O. .2000F+Ol O. .5000E+00 O. O. O. -.2500E+Ol O.
• 2000E+Ol O. .5000E+00 O•

INVE R SE OF STIFFNESS MATRIX


.1 000E-19 - . 1 316E-38 .1930E-20 .1316E-38 .1930E-38 .6579E-57 .3333E-;>0 .7895E-?1 .1226E-J8 -.1316E-38
. 1 404E-20 .1316E-38 .1404E-38 .65 79 E-57
-.1 3 1 6E-38 .1000E-19 -.7895E-21 .2193E-39 -.7895E-39 .5263E-57 .8333E-21 .2632E-20 .1316E-38 .3114E-38
.7895E-21 .1886E-38 .7895E-39 .5263E-57
.1930E-20 -.7895E-21 .6491E-02 .7895E-21 .6491E-20 .3947E-39 .1667E-02 -.5263E-03 .1404E-20 -.7895E-21
.1754E-03 .7895E-21 .1 754E- 21 .3947E- 3 9
.1316E-38 .2193E-39 . 7895E - 21 .100 0E -19 • 7895E- 39 .2 5 0 0E-38 - .6 333E-21 .2368E-20 -.1316E-38 . 1866E-.38
-. 7 895E-21 .6140E-39 - . 7895E-39 .2500E-38
.1 930E-38 -.7895E-39 .6491E-20 .7895E-39 .1000E-19 .3947E-57 .1 667E-20 -.5263E-21 .1404E-38 -.7895E-39
.1 754E-21 .7895E-39 .1754E-39 .3947E-57
. 6579E-57 .5263E-57 .3947E-39 .2500E-38 .3947E-57 .1000E-19 .4815E-53 .1184E-38 -.6579E-57 .5263E-57
-.3947E-39 .2500E-38 -.3947E-57 -.2500E-38
.3333E-20 .8333E-21 .1667E-02 -.8333E-21 .1667E-20 .6525E-53 .4167E-02 .3316E-17 .3333E-20 -.8333E-21
.1667E-02 .8333E-21 .1667E-20 .4366E-53
.7895E-21 .2632E-20 -.5 2 6 3E-03 .2368E-20 -.5263E-21 .1184E-3d .2602E-17 .3421E-02 -.76 95E -21 .2632E-20
.5263E-03 .2368E-20 .5263E-21 .1184E-38
.1228E-38 .1316E-38 .1404E-20 -.1316E-38 • 1404E-38 -.6579E-57 .3333E-20 -.7695E-21 .1000E-Ie) .1316E-38
.1930E-20 -.1316E-38 .1930E-38 -.6579E-57
-.1316E-38 .3114E-38 -.7895E-21 .1886E-38 -.7895E-39 .5263E-57 -.8333E-21 .2632E-20 .1316E-]8 .1000E-19
.7895E-21 .2193E-39 .7895E-39 .526 3 E -57
.1404E-20 .7895E-21 .1754E-03 -.7895E-21 .1754E-21 -.3947E-39 .1667E-02 .5263E-03 .1930E-20 .7895E-21
.6491E-02 -.7895E-21 .6491E-20 -.39 47E-39
.1316E-38 .1886E-38 .7895E-21 .6140E - 39 .7895E-39 .2500E-38 .8333E-21 .2368E-20 -.1316E-]8 .2193E-39
-.7695E-21 .1000E-19 -.7895E-39 .2500E-38
.1404E-38 • 1895E-39 .1754E-21 -.7895E-39 .1754E-39 -.3947E-57 .1667E-20 .5263E-21 .1930 E-38 .7895E-39
.6491E-20 -.7895E-39 .1000E-19 -.3947E-57
.6579E-57 .5263E-57 .3947E-.39 .2500E-.38 .3947E-57 -.2500E-38 .3263 E -5 3 .1184E-38 -.6519E-57 .5263E-57
-.3947E-39 .2500E-38 -.3947E-57 .1000E-19
ITER~TION NUMBER

NODE X-DISPLACEMENT Y-DISPLACEMENT


1 -.000000 .000000
2 .Ot3333 -.000000
3 .000000 -.000000
4 .006667 .000000
5 -.000000 .000000
6 .013333 -.000000
7 .000000 -.000000

SOLID ELEMENT STRESSES

ELEMENT XX-STRESS YY-STRESS XV-STRESS MAJOR PRINCIPAL MINOR PRINCIPAL ORIENTATION

STRAIN DISPLACEMENT MATRIX


-. tOOOE+Ot O. .tOOOE+Ot O. -0. O.
O. -.tOOOE+Ot O. -.tOOOE+Ot O. .2000E+Ot
-.tOOOE+Ot -.tOOOE+Ot -.tOOOE+Ot .tOOOE+Ot .2000E+Ot -0.

STRESS-DISPLACE~ENT TRANSFORMATION MATRIX


-.tOOOE+03 o. .tOOOE+03 O. O. o.
O. -.tOOOE+03 O. -.tOOOE+03 O. .2000E+0]
-.5000E+02 -.50COE+02 -.5000E+02 .5000E+02 .tOOOE+03 O.
-.3667E+Ot .9715E-t4 .2842E-t3 .7105E-14 -.3667E+Ot 90.t6

STRAIN DISPLACEMENT MATRIX


.tOOOE+Ot O. .1000E+(jt O. -.2000E+Ot O•
O. -.tOOOE+Ot O. • 1 OOOE+O 1 O. O.
-.tOOOE+Ol .tOOOE+Ot .tOOOE+Ot .tOOOE+Ol O. -.2000E+Ol

STRESS-DISPLACEMENT TRANSFORMATION MATRIX


.tOOOE+OJ O. .tOOOE+03 O. -.2000E+03 O.
O. -.tOOOE+03 O. .tOOOE+03 O. O.
-.5000E+02 .5000E+02 .5000E+02 .5000E+02 O. -.tOOOE+03
2 -.3667E+Ot .4815E-32 .2356E-13 O. -.3667E+Ot 90.t6

STRAIN DISPLACEMENT MATRIx


.tOOOE+Ot O. -.tOOOE+Ot O. -0. O.
O. .tOOOE+Ot O. .tOOOE+Ot O. -.2000E+Ol
.tOOOE+Ot • tOOOE+Ol .tOOOE+Ot -.tOOOE+Ot -.2000E+Ot -0 •

STRESS-DISPLACEMENT TRANSFORMATION MATRIX


.tOOOE+OJ O. -.tOOOE+03 O. o. O.
O. .tOOOE+03 O. .tOOOE+OJ O. -.2000E+03
.5000E+02 .5000E+02 .5000E+02 -.5000E+02 -. tOOOE+03 O.
3 -.3667E+Ot -.97t5E-t4 O. -.7l05E-t4 -.3667E+Ot 90.t6

STRAIN DISPLACEMENT MATRIX


-. tOOOE+Ot O. -.tOOOE+Ot O. • 200.0E +0 1 O.
O. .tOOOE+Ol O. -.tOOOE+Ol O. O.
• tOOOE+Ot -.tOOOE+Ot -.tOOOE+Ot -.tOOOE+Ot o• .2000E+Ot

STRESS-DISPLACEMENT TRANSFORMATION MATRIX


-.tOOOE+03 O. -.tOOOE+03 O. .2000E+03 O.
O. .tOOOE+03 O. -.tOOOE+03 o• O.
• SOOOE+02 -.5000E+02 -.5000E+02 -.5000E+02 o. .tOOOE+03
4 -.3667E+Ot -.60t9E-32 .4857E-t4 O. -.3667E+Ot 90.t6
JOINT DtSPLACE~ENTS JOI NT STRESSES
h
ELEMENT SHEAR NORMAL ROTATION SHEAR NORMAL fOIOMENT
:g
CD
JOINT DISPLACEMENT VECTOR - LOCAL COORDINATES ;::,
-.1566E-52 -.1833E-19 -.5220E-53 -.lB33E-19 -.3947E-39 -.1333E-01 -.4815E-3~ -.1333E-01 ct
5 -.2~07E-34 -.1333E-Ol -.61Q6E-15 -.2407E-32 -.3667E+Ol -.3053E-13 )('

UP= -.4666E-Ol UR= -.1866E+00 TORP= .4666E+Ol TORR= .2868E+Ol TOR~ -.2401E-32
OILAT= O. DELV= -.1333E-Ol

UP= -.4666E-Ol UR= -.1866E+00 TORP= .~666E+Ol TORR= .2868E+Ol TOR= -.2407E-32
OILAT= O. DELV= -.1333E-01

INCREMENTAL FORCES FOR JOINT


-0. .9697E+00 -0. .9697E+OO O. -.9697E+00 O. -.9697E+00

JOINT STIFFNESSES
O. O. O. O. O. O. O. O. .1000E+03 .2000E+03

INITIAL STRESSES AT END OF INCREMENT


SHEAR NORMAL MOMENT SHEAR NORMAL MOMENT
-.5000E+OI -0. -0. -.5000E+Ol -0. -0.
-.5000E+~1 -0. -0. -.5000E+Ol -0. -0.
O. -.2939E+Cl -.2665E-13

INCREMENTAL LOAD VECTOR


O. O. -.9697E+00 O. • Q697E+00 O. O. o. O. O•
-.9697E+00 O. .9691E+00 O.

fg
~
394 Appendix 1

LISTING OF FINITE ELEMENT PROGRAM - - "JETTY"

PPOGRAM JETTY(I~PUT,OUTPUTpTAPE5=INPUT,TAPEo=OUTPUT,TAPE7)
c
C PPOGRAM FOR ANALYSIS OF JOINTED ROCK STRUCTURES USING THE FINITE
C ELEMENT MET~OD WRITTEN ~V OR. C. ST. JOHN WITH T~E COLLABORATION
C OF DR R F GOnD~AN, USING THE THEORETICAL DISCUSSIO~ OF CHAPTER B
C IT IS A SMALL PROG~AM WITH LIMITED HORIZONS. ITS PRINCIPAL
C PURPOSE IS FDUCATIONAL, I.E. T~ Df.MONSTRATF. HOW THF NUME~ICAL
C MODEL CAN BE CODED IN PRACTISE. THE EQUATIONS REFERRED TO IN
C THE NUMEROUS COMME~T CARDS H~REIN ARE EQUATIONS OF CHAPTE~ 8
C
C LIST OF ~AIN ARRAVS AND THEIR USES
C
C MTVPE= ELEMENT MATERIAL OR JOINT TVPE
C KS GLOBAL STIFFNESS MATRIX
C R VECTOR OF NODAL FORCES
C ~XV 2*N MATRIX EQUIVAL~NT TO R - USED FOR INPUT ANO OUTPUT
C U VECTO~ OF NODAL DISPLACEMENTS - useD JNITIALV FOR WATER FORCE
c: UXV 2*N MATRIX ~aUIVALENT TO U - USED FOR INPUT AND OUTPUT
C X VECTOR ~= X-ORDINATES OF NOnES
C V VECTOR OF V-ORDINATES OF ~ODES
C ICODE= VECTOR OF DISPLACEMFNT CO~DITIONS
C 1 - FIXED IN X DIRFCTION
C 2 - FIXED IN V DIRECTION
C 3 - FIXED IN X AND V DIRECTION
C NOD N*4 MATPIX OF NODES n~FINING ELEMENTS
C RESID= N*3 MATRIX ~F RESIDUAL STRESSES
C SIGNO= N*3 MATRIX OF INITIAL STRESSES
C C GLOBAL STRESS STRAIN MATRIX FOR SOLID ELEMENT PROPERTIES
C ALSO USE~ TO STORE JOINT ELEMENT ~ROPERTIES
C RJ VFCTOR OF MASS DENSITY OF DIFFE~ENT MATERIAL TYPES
C G SFT TO ZEPO IF GRAVITV NOT ACTIVE
C SET TO ONE IF GRAVITV ACTIVE
C KJS VECTOR OF JOINT SHEAR STIFF~ESSFS
C KJN VECTOR OF JOINT NORMAL STIFFNFSSES
C KJC; NU= NEW JOINT SHEAR STIFFNESSF.S
C KJNNLJ= NFW JOINT NORMAL STIFFNESSES
C ITFP1= NU 'IIIBER OF FIRST ITERATION THIS RUN
C ITeRN= NU"4RER OF LAST ITFRATION THIS RUN
C NOOF.S= NUMBER OF NODAL POINTS
C NCLEM= NUMBFR OF ELEMENTS
C N30UN= NUMBER OF NODES WITH DISPLACEMENT CONSTRAINTS
C NSDL = NUMRER OF SOLID ELFMENT TVPF.S
C JOINT= NUMBER OF JOINT ~LfMENT TYPES
C ITER = NUMBER OF ITERATIONS THIS RUN
C G~AV = ACCELERATION DUE TO GRAVITV
C ACCf"L= ACCELERATION NOT DUF TO GRAVITY - IN G
C ~IR DIRECTION OF ACCEL - + ANTI CLOCKWISE FROM HORllONTAL
C L~AD SFT TO ZERO. 8EC00.4[S 1 IF AN ERROR IS CNCOUNTJ::RED
C NUM VECTOR OF NODES WITH nISPLAC~MENT CONSTRAINTS
C HEAD 1.0 FOR SOLID ELEMENT ABOVE WAT~R TABLE
C 0.0 FOP SOLID ELEMeNT AELOW WATER T~9LF
C
COMrJlON /ME.:SH/ X(25) ,Y(25) ,R(50) ,U(5()) ,KS(50,50),
1 NOD(23,4),PESID(23,1),StGNO(23,3',MTVPf(23),
2 NODES,NELEM,NSOL,NFREE
Rf~L KS

C 0\1 M0 N / C; T F F / K J S ( 2 3 ) , K J N ( 2 3 , , K J S NU ( 2 3 ) , KJ "I NU ( 23 ) , I T E R 1 , I T E ~ N , I T
REAL KJS,KJN,KJSNU,KJNNU
COMMON /BOUN/ NUM(20),ICOOE(20),NBOUN
Appendix 1 395

COM"ION /(ON5/ RO(10),A(ELX,A(ELV,NRES,INTER,(3,3,10)


D I ~E NS I ON IDA T ( 131 , I (A R D ( 9 I , E ( 3, 3 ) , T ( 3 , .3 ) ,H ( 3,6 I ,
1 RXV(2,25),UXV(2,251,CJ(9,lC)
~QUIVALfNCE (R(l),RXV(l,l»),(U(1 ),UXYCl,l»),(CJ(I,l),(l,l,l»

DATA 11AT /RHNnD~L PO,AHELF~ENT ,~HHVOqAULI,8HRESIDUAL,eHBOUNDARV,

18HSDLI~ EL,8HJOINT EL,BHSTART ,~HSTOO ,8HRESTART ,8HINTER~ED,


28HA( ( FLFRA,~HSAVE ST[/

C
DP"ENSION 8(5~), SCRTCH(lSO)
( THESE ARE NEFDcn FORTHE MATRIX INVERSION SU9ROUTINE
C
C READ THE TITLE CARD AND THEN WRIT~ IT
C
R F'&'O C 5, lOr:: ': 1 ICARD
~RITE(6,2:)O,,) If:APD
c
C READ ESSENTIAL CONTROL DATA AND wRITE IT OUT I~MEDIATELV

C
RF.ADC5,10~l' NODES,NELF~,N~OUN,NSOL,JOINT,ITERl,ITCR,G,GRAV,ACCEL,
1 DIR,PW
IF CITsRl . f O. 01 ITFRl:: 1
WRITE(~,2001)NQDE5,N~LEM,N90UN,~50L,JOINT,ITER1,ITER,G,GRAV,ACCEL,
1 DIP,RW
ITERN = [TERl+ITFR-l
C
C S~T SO~E INITIAL VALues
C
NTQT= NSOL+JOINT
NR~S=O
LBAD=O
INTEP =
PI=ATAN(1.C'*4.G
C
C SET 50"'= OF THE ~qRAVS TO ZERO BFFORE STARTING
(
NFREE=NODES+NODES
(
DO 1 1=1,50
RCf)=O.O
LJ(I)=C. ':
CONTINUE
DO 2 I=I,NELEM
KJS(I)=O.O
KJN( 1)::0.0
KJNNU( I )=0.'2
KJ SNU C I ) ::(). 0
00 2 j =1 , 3
RE SID C I , j ) :: J • C
5 I GNO ( I , J) = 0 • C
2 (OPI.ITINUE
C
C NO~ CALCULAT~ TOTAL ACCELERATION COMPONENTS - EQUATI9N 66A IN r~XT

C
DIR=DIR*PI/180.
ACELX=ACCEL*COS(OIR).GRAV
C
C G C~N TAKE VALU~ 1.0 OR 0.0 CORRESPONDING TO GRAVITV ON OR OFF
C
ACELV=I-G+ACCF.L*SIN(OIR»*GRAV
C
C q~AD NEXT OAT~ CARD -THIS MUST BE A HEADER CARD FOR A OATA SET
C OP A START / STOP / RESTART / INTER
396 Appendix 1

100 READ (5,1000) ICARD


C
C DETER~INE WHICH DATA S~T FOLLOWS
C
J=ICARO(l)
DO 130 1=1 , 13
IF(J.EQ.IDAT(Y)' GO TO 150
130 CONTINUE

~RITE (6,2002) ICARD


LAAD=LBAO+l
GO TO 100

C
CALL INTERMEOIATE RESULTS DESIRED ON PRINTOUT
C
SINTER = 0
GO TO 10C
c
C RFAD NOO~L POINT INFORMATION
C
1~ L=I)
11 READCS,1004) N,X(N' ,V(N) ,RXV( 1 ,N) ,RXY(2,N,
c
C A SIMPLE NODAL POINT GENERATOQ AND ~RROR SFARCH
C
IF(L.Ea.o, GO TO 12
lX=N-L
DY=(YCN)-Y(L,,/ZX
OX=(X(N)-X(L')/ZX
12 L =L + 1
IF- (N ... L' 15,14,13
13 X(L)=X(L-I,+DX
V(L)=Y(L-l)+DY
GO TO 12
14 IF (NODES - N) 15,16,11
C
C NOOAL POINT DATA ERROR WQITE EqROR MESSAGE
C
15 WR I TE ( 6,20 I) 4' N
Lt;!AD=LBAD+l
Jf(N.LT.NODES' GO TO 11
C
C WRITE OUT NODE OATA
c::
16 WRITE(6,2003) (K,X(K),YCK),RXY(1,K),RXY(2,K),K=I,NODES)
GO TO 100
c
C F.~EMENT DATA - MATERIAL TYPE ANO NODAL POINTS
C
20 N.O
21 READ(S,100S) K,MAT,(NOD(K,I),I=I,4)
C
C A SIMPLE FLEMENT GENERATOR ANO F.RROR SEARCH
C
MTYPF,(K)=MAT
IF(MAT.GT.NTOT' GO TO 27
22 N=N+l
IF CK - N'27,26,23
23 Nl=N-l
MAT:;: MTYPE(Nl'
Append~ 1 397

MTVPE(N)=MAT
NOD(N,J )=NOD(Nl,l'+J
NOQ(N,2,=NOOCN1,2)+1
NOD(N,3)=NODCN1,3)+1
IF CMAT .GT. NSOL) GO TO 24
NOOCN,4)=O
GO TO 25
24 NOD(N,4)=NQD(Nl,4)+1
25 IF CK - N) 27,26,22
26 IF (NFLEM - N) 27,29,21
c
C EL~MENT ~ATA ERRQR - WRITE ERROR MESSAGE
C
27 WRITE(6,2007) K
lBAO=LBAQ+l
IF(K.LT.NELEM) GO TO 21
c
C ~PITF OUT FLEMENT DATA
C
~~ WRITE(6,?C~6) (K,MTVPECK',(NOD(K, 11,1=1,4),K=J ,NELEM'
GO TO 100
c
C READ HYORAULIC GRADIFNT DATA FOR ELEMENTS
C
30 L=O
31 READ(5,1004) N,HEAD,DHDX,OHDV
c
C ZERO ENTRIES FQR JOINTS AS RESIDUAL STRESSES WILL BE EFFECTIVE
C FOR A SOLID ELEMENT HEAD =1.0 IF BELOW WATER TABLE
c HEAD = 0.0 IF ABOVE WATER TABLE
c
C
C CALClJLATE AND THEN STORE TEPMS REQUIRED FOR EQUATION 6S
C PRESSURE GRADIENTS ARE STORED TO AVOID HAVING TO STORE
C THE WATER TABLE SWITCH
C
GW=GRAV*HEAD*RW
UXYC1,N)=DHDX*GW
UXVC2,N)=(DHDV-l.0'*GW
c
C GENERATE MISSING DATA AND CHECK FOR ERRORS
C
32 L=L+l
IF (N - L) 35,34,33
33 Ll=L-l
UXVC1,L)=UXVC1,Ll'
UXV(2,L)=UXY(2,Ll)
GO TO 32
34 IF (NELFM - N, 35,36,31
C
C HVDRAULIC GRADIENT CARD ERROR - WRITE MESSAGE
C
35 WRITE(6,2010) N
LBAD=LRAD+l
IF(N.LT.NELEM) GO TO31
C
C WRITE OUT PRESSURE GRADIENTS FOR ELEMENTS
C
36 WRITF(6,2009) (K,UXVC1,K),UXV(2.K),K=1,~LEM)
GO TO 100
c
C READ n~SIDUAL STRESS CAROS
C NRES SET T, 1 TO INDICATE THAT ~ESI)UAL STRESSES HAVE BEEN READ
398 Appendix 1

C
40 NRES;;I
L=C
41 READ(S,1004' N,RESI~(N,I"RESID(N,2"RESIDrN,3)
C
C GENERATE MISSING DATA AND THEN CHECK FOR ERRORS
C
42 L=L+I
IF (N - L, 4S,44,43
43 Ll;;:;L-1
RESID(L,l '=RESIO(Ll.l,
RESID(L,2,=RESID(LI,2)
RESID(L,3)=RESID(Ll,3)
GO TO 42
44 IF (NELEM - N, 4S,46,41
C
C RESIDUAL STRESS CARD ERROR - wRITE ERROR MESSAGE
C
4S WRITE(6,2013) N
LBAD=LBAD+l
IF(N.LT.NELEM) GO TO 41
C
C WRITE OUT RESIDUAL STRESS DATA
C
46 WRITE(6,2012) (K.RESID(K,I'.RESIO(K,2',RESID(K']',K=I,NFLEM'
GO TO 100
C
C READ LIST OF NODAL CONSTRAINTS AND THEN WRITE THEM OUT
C
SO READ(S,IOOS) (NUM(K),ICO~E(K),K=I,NBOUN)
WRITF(6,20IS) (NUM(K),ICODE(K),K=I,NBOUN'
GO TO 100
C
C REAO SOLID ELEMENT PROPERTIES AN~ THEN WRITE THEM
C
60 WRITE(6,~017)
00 69 K=I,NSOL
READ(S,1004) MAT,RO(MAT),ES,EN,GSN,PRSN,PRST,ALPHA
WRITE(6,2018~MAT,RO(MAT),ES,EN,GSN,PRSN,PRSTtALPHA
C
C F~~~ STR~SS-STRAIN MATRIX IN LOCAL CO-ORDINATES - EQUATION 208
C
RAT=EN/ES
CONST=1.C/«1.0+PRST'*(1.0-PRST-2.0*RAT*PRSN.PRS~))
E( 1 ,I ,=(ES-EN*PRSN*PRSN)*CONST
E(2,1)=EN*PRSN*(1.O+PRST)*CONST
F.(3,1)= 0.0
E(l,2'= E(2,1)
E(2,2'=EN*(1.0-PRST*PRST'*CONST
E(3,2'= 0.0
E( 1 ,3'= 0.0
f(2,])= 0.0
E(3,3)=GSN
C
IF (INTER .EQ. 0) WRITE (6,99Q) «E(J,L),L=I,3),J=I,31
999 FOQMAT (44Ha STRESS-STRAIN MATRIX IN LOCAL CO-ORDINATES/
I (SX,3E15.S"
C
C NOw FORM THE TRANSFORMATION MAT~IX - EQUATION 23.
C
ALPHA=ALPHA*PI/IBO.C
COSA=COS(ALPHA)
SINA=SIN(ALPHA)
A ppendix 1 399

T ( 1 ,1 ) =C OS A*C OSA
1( 2,1 1= S I NA*S IN A
T( 3,1 )= - 2 . 0 * SIN A* COSA
T (I, 2)= T(2 ,1)
T( 2, 2)= 1 (1 , 1 )
T( 3 , 2 ) =- T ( 3, 1)
T( I,3 ) -= S INA*COSA
T( 2,3 ) = -T(1,3)
T (3,3)= T ( l ,I)-T( 2, 1)
C
IF (I NT ER . E'O. 0) WRfTf" (0,9 98) «(T( J, L ) ,L=1,3),J=1,3)
999 FO QMAT (23 HC T RA NSFO QMA TI ON MATRIX/(5X, 3E 15.5»
C
C NOW T~ANSF ORM TO GLO BAL CO-ORD INA L SCHEME - US E H AS TFMP. STORAGE'
C
DD 66 N= I ,3
DO 66 J -:l ,3
He J,N) = C. C
DO 6 6 L=I ,3
H (J,N) =H(J,N)+E (J,L )*T ( L ,N)
66 CONTI NUE
DO 6 7 N=I,3
DO 67 J= 1 , 3
CeJ ,N,MA T )= O.O
00 67 L=1 , 3
C( J,N,MA T )=CCJ,N,MA T)+T(L,J )*H( L,N)
67 C O ~T rN UE

C
I F eINTER .CO. 0) WRITF e6,997) « C ( J, N, MA T),N-=I, 3) ,J=I,3)
997 FORM AT (44 H ~ STRE SS -ST RA I N MA TRIX I N GLOBAL COOR DI NATES/
J ( SX,3 EI5. S »
(
C (HE(K FOR ERRORS
C
IF (N S OL .GF. MAT ) GO TO 69
WRI TE(6, 2 01 Q)
LBAD =L BAD+l
69 CON T IN UE
GO TO 100
C
( RE AD J OINT PROPCqTICS AND STORF IN C
(
70 WR I TE(6,2 02C' )
NT OT = NSOL+JOINT
DO 75 J=I, J O INT
REAO(S,100 4 ) MAT ,(C J( K, MA T),K=1,8)
(
C C O~PRFSSIVE STREN GTH NEGATIVE AN~ TENSILE TO COMPPESSIVF RATIO +VE
( MAXIMU M CLOSU RE MUST RE POSITIVE
C S E AT I NG LO AD MUST ALW AYS BE NEGATIVE
C
CJ( l, MA T)=-A AS( CJ ( l ,MAT»
C J (2 ,MAT )=ABSe CJC2,MA T»
CJ(5,~AT )=AB S(CJ (5 t MA T')
(J (6 , MAT'=-ABS( C J ( 6,~AT)'
(
WR IT ::(6,2021 ) MA T, (CJ(K,M~T) , K =l ,R)
c
C CHE CK FO R FRROR S
r
IF ( NTO T .GF:. MAT) GO TO 75
WRI TE (6, 2022)
LBAD=LBAO+l
400 Appendix 1

75 CONTINUE
C
C IF THIS NOT THE FIRST RUN THEN JOI~T STIFFNESS IS TO BE REA~
C
IF(ITERI.GT.I) READ(?) KJS,KJN
GO TO l~O
C
C DATA INPUT COMPLETE - PROCEED IF NO ERRORS
C
BO IF (LBAD .EO. 0) GO TO 300
C
C ERRORS HAVE BEEN FOUND - WRITE MESS_GE
C
WRITE(6,2023)
STOP
C
C RESTART 8Y READING OATA FqOM TAPE?
C
95 RE'AD(7)
GO TO ~75
C
C CALL SUAROUTINE STIFF TO ASSEMBLE THE STIFFNESS MATRIX
C
300 CALL STIFF
C
C CALL LIBRARY SUAROUTINE TO INVERT THE STIFFN~SS MATRIX
C ~FRE~ IS THE SIZE OF THE STTFFNESS MATRIX
C IN LAPGER PROGRAMS, STIFFNESS MATRIX SHOULD ~E NARROW BANDED
C
IF (INTER .NE. 0) GO TO 350
WRITt=:(6,996)
996 FORMAT ()4HO LOAD VECTOR AND STIFFNESS MATRIX)
DO 995 I=l,NFREE
WRITE(6,994) (KS(I,J),J=l,NFREE)
994 FORMAT(2X,10E12.4)
9q5 CONTI NUE
WRITE (6,994) (R(I),I=l,NFREE)
C
350 CALL LINV3F (KS,A,1,NFREE,50,-1.0,02,SCRTCH,IER)
C
C KS IS NOW THE INVERSE OF THE STIFFNESS MATRIX
C
C
IF (INTER .NE. 0) GO TO 359
WRITE (6,993)
993 FORMAT (29HC INVERSE OF STIFFNESS MATRIX)
DO 992 I=l,NFREE
992 WRITE (6,994) (KS(I,J),J=I,NFREE)
C
C SET INITIAL STRESSES TO RESIDUAL STqESSES RFFORE STARTING
C
359 DO 360 M=l,NELEM
SIGND(M,l)=RESID(M,l)
SIGND(M,2)=RESTD(M,2)
STGNO(M,3)=RESIO(M,3)
360 CONTINUE
C
C SFT INITIAL DISPL~CEMENTS TO lERO
C
DO 37C J=I,NFRFF
U(J)=O.O
370 CONTINUE
c
Appendix 1 401

C ~OW ENTER MAIN ITeRATIVE LOOP


C
375 DO 500 IT=ITER1,ITERN
WRITE(6,2026) IT
C
C DISPLACEMENTS ARE ACCUMULATED - EACH ITERATION GIVES NEW INCRE~ENT
C DISPLACEMENT VECTOR (U)= U + KS • LOAD VECTOR
C
DO 400 J=l,NFREF
DO 400 K=l,NFREE
U(J)=U(J)+KS(J,KI*R(K)

400 CONTINUE
C
C WRITE OUT NODAL DISPLACEMENTS
C
WRITF.(6,20?4) (J,UX,( 1 ,J) ,U)(Y(2,J) ,J=l ,NODES)
C
C SfT LOA~ V~CTOR TO ZERO FOR INCRE~ENTAL LOADS
C
DO 450 J=l,NFRFE
R(J)=O.O
450 CONTINUE
C
C CALL STRFS TO CALCULATE SOLID ELEMENT STRESSeS
c
CALL STRES
C
C CALL JSTR TO CALCULATF JOINT 9EHAVIOUR AND NEW LOAD VECTOR
C
IF(JOINT.GT.O) CALL JSTR
C
5eo CONTINUE
C
GO TO 100
C
C ACC=:LE'RATE WRITE NEW JOINT STIFFN=SSFS ONTO TAPE
c
600 WRITE (7) KJSNU,KJNNU
GO Tn 700
c
C SAV~ STIFFNESS WRITE ALL DATA AND INvERTFD STIFFNESS MATRIX ON TAPE
C
6S0 WRITE (7) C,X,'(,R,U,KS,NUM,NOD,RESID,SIGNO,MT,(P~,ICODE,KJS,KJN

c
7C::} STOP
C
C INPUT FORMAT STATEMENTS
C
1000 FORMAT(9A8)
IO~1 FORMAT(7IS/SFIC.OI
1004 FORMAT(IS,7FIO.O,F5.0)
1005 FORMATCI615)
C
C OUTPUT FORMAT STAT~MENTS
C
2000 FORMAT(IHl,lOX,9A8)
2001 FORMAT( 30HO NUMBER OF NODAL POINTS =,13 /
1 30H NUMBER OF ELEMENTS ::, 13 /
2 30H NUMAER OF CONSTRAINED NODES=,I) /
3 30H NUMB~R OF SOLID MATERIALS =,13 /
4 30H NUMBER OF JOINT TYPES ::, 13 /
30H NUMOER OF FIRST ITERATION =,13 /
402 Appendix 1

5 30H NUMBER OF ITERATION~


6 30H GRAVITY LOADING - ON/OFF =,F4.r. /
7 30H ACCELERATION DUE TO GRAVITY=,CI2.4 /
8 30H PSEUOO STATIC ACCELERATION =,EI2.4 /
9 30H DIRECTION OF ACCELERATION =,EI2.4 /
A 30H MASS DENSITY OF WATER =,fI2.4 )
2002 FORM_T C30HO FAULTY INPUT CARO CONTAINS 9A8)
20e3 FORMAT (IH02C3X,4HNODE6X,4HXORD8X,4HYORD7X,5HXLOAD7X,SHYLOAD5X)/
(2CI7,2F12.2,2EI2.4,3X»)
2004 FORMAT (19HO DATA ERROR - NOOEI5)
~OC6 FORMAT (lH02(5X,7HELEMENT4X,AHM~T TYPE7X,9HNODE LIST7X)/
1 (2(Ill,110,5X,415»))
2007 FORMAT C22HO DATA EqROR - ELEMENT 15)
2009 FOR"'AT (IH02C2X,7HfLEMENT7X,5HOP/DX6X,10HDP/DY OR P)/
C2CI9,2EI4.4» )
2()10 FORMAT (47H0 DATA ERROR - HYDRAULIC GRADIENT CARD, ELEMENTIS)
2012 FORMAT (lH02(1X,7HELEMfNT9X,29HRESIDU~L STRF.SSES XX,VY,XY7X)/
1 (2(II0,3E1S.4»)
2013 FORMAT (44HO DATA ERROR - RESIDUAL STRfSS CARD, ELEMENT IS)
2015 FORMAT (IH05X,23HNOOE DISPLACEMENT CODf/(19,II5»
2017 FORMAT(IHC,2X,2SHSDLID MATERIAL PROPERTIES)
2018 FOq~AT( 2X,4GHMATERIAL TYPE =,13 /
1 2X,40HMASS DENSITY =,EI2.4 /
2 2X,40HYOUNGS MODULUS - PARALLEL TO ANTSOTROPY=,EI2.4 /
3 2X,4~HYOUNGS MODULUS - NORMAL TO ANISOTROPY =,E12.4 /
+ ?X,40HSHEAR MODULUS - GSN =,EI2.4 /
4 2X,40HPOISSONS RATIO - SN =,F8.4 /
5 2X,40HPOISSONS qATIO - ST =,F~.4 /
7 2X,41HORIENTATIUN OF LAYERS (ANGLE S-X) =,FB.4)
2019 FORMAT (37HO DATA ERROR - MATERIAL PROPFRTY CARD)
2020 FORMAT(lHO,20H JOINT OROPERTIES )
2021 FORMATC lX,41HJOlNT TYPE NUM9ER =;13 /
1 lX,41HWALL POCK COMPRESSIVE STRENGTH =,E12.4 /
2 lX,41HRATIO OF TENSILE TO caMPR~SSIVE STRENGTH=,F8.4 /
3 lX,41HSH~AR STTFFNESS =,EI2.4 /
4 lX,41HRATIO OF R~SIOU~L TO PEAK SHEAR STRENGTH:,E12.4 /
0;, lX,41HMAXIMUM NC)RMAL CLOSURE =,EI2.4 /
V lX,41HSEATING LOAD =,EI2.4 /
6 lX,41HFRICTION ANGLE OF A SMOOTH JOINT =,FI2.4 /
7 lX,41HDILATANCY ANGLE =EI2.4/)
2022 FORMAT (34HO DATA ERROR - JOINT PROPERTY C4RD)
2023 FOQMAT (41H0 DATA ERRORS ~ETECTED - PROGRAMME H4LTEO)
202~ FORMAT (44HO NOOE X-DISPLACEM~NT Y-DTSPLACFMENT/
1 (I6,2FI9.6)
2026 FORMAT (I~Hl ITERATION NUMBE~IS//)
C
END

SUAROUTINE STIFF
C
C SUAROUTINE TO ASSFMBLF MODeL STIFFN~SS ~ATRIX AND LOAD VfCTOR
C AT ENO OF SUARDUTTNE THE MATRIX IS READY FOR INVFRSION
C
rOMMON /MESH/ X(2S),Y(2S),R(50),U(SO),KS(SC,5C),
1 NOD(23,4),RESID(23,3),SIGNO(23,3),MTYPE(23),
2 NODES,NELFM,NSOL,NFREE
RFAL KS
COMMON /BOUN/ NU~(20),ICODE(20),NAOUN
COMMON /FORC/ KSUR(8t8),F(8),LBADt~,H(Jt6)
REAL KSUA
Appendix 1 403

DIMENSION RFS(3),XT(4),YT(4),LAB(4)
C
C SET STIFF NE SS KS TO ZERO
C
LAAD = 0
00 100 J=l,NFREE
DO 1~0 l=l,NFREE
KS(I,J)=O.O
100 CONTINUE
C
C FOR EACH ELEMENT CALCULATE CONTRIAUTION TO GLOBAL STIFFNESS AND LOADS
C
DO 450 M=I,NELEM
MAT=MTYPF.fM)
C
C 15 !T ~ JOINT OR A TRIANGULAR ELEMENT
C
NUMB = 3
!F (MAT .Gr. NSr)L) NUMR 4
C
C EXTRACT NODAL COORDINATES FROM NODAL COOR~IN~TE V~CTnRS

C
250 l=l,NUMB
0(1
N="!OO (M, I )
XT( II=X(N)
YT(I)=Y(N)
250 CONTINU E
C
C SEL E CT JOINT OR TRIANGULAR STIFFN~SS SUAROUTINE
C
M2=M+M
PES(I)=RE510(M,1)
RES(2)=RESIOfM,2)
RES(3)=RESID(M,3)
IF (MAT .GT. NSOLl GO TO 270
c
[)POX=U(M2-1)
OPf)Y=U(M2)
CALL TRJAfO,MAT,XT,YT,RFS,DPDX,~PDY)
IF (LAAO .NE. 01 450, 28(')
c
27C CALL JSTIF(MAT,XT,YT,RFS)
IF (LBAD .GT. 0) GO TO 450
C
C RETAIN TRANSF~RMED RESIDUAL STRESSES
C
PE510(M,1)=RES(11
RESIDfM,2)=RES(21
RESIDfM,3)=R~S(3)

C
C NOW ADD THE ELEMENT CONTRIBUTIONS KSUO AND F TO KS AND R
C ASSFMALE VECTOR INDICAT[NG LOCATION OF ELEMfNT CONTR!BUTION
C
280 0(1 300 !=l,NUMB
J=NOD(M,I)
LAA([J=J+J-l
300 CONTINUE
C
00 400 J=l,NUMB
[I=LAEHII
[K=[+f-l
12=1 [+1
C
Goodman-Geologic al Eng .- 2·7
404 Appendix 1

C LOAD CONTRIBUTION
C
R( Ir)=R(II)+F(IK)
R( I2)=R(I2)+F( IK+l)
DO 400 J=I,NUMA
JJ=LAB(J)
JK=J+J-l
J2=JJ+l
C
C STIFFNESS CONTRIBUTION
C
KS(II,JJ)=I<S(II,JJ)+"'SUFI(IK ,J'"
KS( II ,J2)=KS(I I ,J?)+KSUA( IK ,JK+l)
KS(12,JJ)=KS(12,JJ)+KSUA(IK+l,JK
KS(I2,J2)~KS(12,J2)+KSUB(IK+l,JK+l)

4CO CONTINUE
450 CONTINUE
IF (LAAO .GT. 0) GO TO 600
C
C NOW INTRODUCE NODAL CONSTRAINTS
C
C ICDDE - FIXED IN X DIRECTION
C ICODE 2 - FIXED IN Y DIRECTION
C ICODE 3 - FIXED IN X AND Y DIRECTION
C
C SET DIAGONAL TERM TO VERY HIGH STIFFNESS
C EFFECTIVELY FIXES NODE IN REOUIRED DIRECTION
C FOR GIVEN DISPLACEMFNTS SET R(JJ)=DISPLACEMENT*1.F+20
C
DO 500 I=I,NBOUN
L=ICODE(I)
IF(L.EO.O) GO TO 500
J=NUM( I)
JJ = J + J
IF (L .EO. 2) GO TO 470
KS(JJ-l,JJ-l) = KS(JJ-I,JJ-l) + 1.0E20
IF (L .EO. 1) GO TO 500
470 KS(JJ,JJ) = KS(JJ,JJ) + I.OE20
500 CONTINUE
C
C
PETURN
C
C ER~ORS DETECTED
C
600 WRITE(6,2000)
2000 FORMAT (46HC PROGRAM STOPPED AS ERRORS HAVE HF.FN DETECTED)
STOP
C
END

SUAROUfINE TRIA(KUT,MAT,X,Y,RESID,DPDX,DPDY)
c
COMMON /CONS/ PO(lO),ACFLX,ACELY,NRES,JNTER,C(3,3,lO)
COMMON /FORC/ KSU8(B,8),F(8),LAAD,M,H(3,6)
PEAL KSUB
DIME~SION ~(4),Y(4),RFSID(3)
REAL LNORT(3,6)
C
Appendix 1 405

C SUBROUTINE TO CALCULAT~ STIFFNESS MATRIXCKSUB) OF TRIANGULAR ELEMENT


C TOG~THER wITH CONTRIRUTION Tn LOAD VECTOR FROM RFSIDUAL STRESSES,
C GRAVITY LOADING, ACCELERATION LOADING AND WATER FORCES (F'
C
C IF <UT IS EQUAL TO 1 THIS ROUTINE WAS CALLED FqOM SUAROUTINE STRESS
C TO CALCULATE THE STRESS-DISPLACEMENT TRANSFORMATION MATRIX (H) QNLY
C
C START BY SHIFTING ORIGIN SO THAT LOCAL COORDINATES FOR FlqST
C NODE AR~ 0.0,0.0
C
XJ=X(2)-X(1)
YJ:Y(2)-Y(1)
XK=X(3)-XCl)
YK=Y(3)-YCl)
c
C NOW CALCULATF TWICE THE AREA OF THE TRIANGLF
C

C
C IF THIS AREA IS ZERO OR NEGATIVE THERE IS AN ERROR IN THE DATA
C RFTUPN IF AN ERROR HAS BEEN FOUND
C
tF(AREA.GT.O.O) GO TQ 10
Lf3AD=LBAD+l
WRITEC 6 ,200G ' LB AD ,M
2000 FORMAT (7HO EPRORI3,9H, EL~MENTI5,22H HAS ~ON-POSJTIVE AREA)
RETURN
C
C FORM STRAIN DISPLACEMENT MATRIX -LNORT -EQU AT ION 16 IN TEXT
C
C NOW CALCULATE THE NECESSARY TERMS - TOGETHER wITH THEIR SIGNS
C
1~ XK=-XK/AREA
YK= YK/AREA
XJ= XJ/ARt:A
YJ=-YJ/AREA
XK J=-XK -XJ
YJK=-YJ-YK
C
C SET LNORT AS EQUATION 16
C
LN()RT(l,I)= YJK
LNORT(2,I'= o.~
LNORT(3,1)= XKJ
LNORTCl,2)= o.n
LNORT(2,2)= Xt<'J
LNORT(3,2'= YJK
LNORT(I,3)= YK
LN.)RT(2,3)= 0.0
LNORT(3,3'= XK
LNDRT(1,4'= 0.0
LNClRT(2,4,= XK
LNORT(J,41= YK
LNORT(1,5)= YJ
LNORT(2,S'= 0.0
LNORT(J,5'= XJ
LNORT(1,6'= 0.0
LN;)RT(2,6,= XJ
L NORT ( J,f>,= YJ
C
IF (I N T E R • Ea. (') WR I T EO (6 , Q Q 9 , ( ( L NO R T ( t , J) , J =I ,6' , I :: t ,3 )
99~ FORMAT (28HO STRAIN DISPLAC~MENT ~ATRIX/(5X,6E12.4»
C
406 Appendix 1

c:
C NOW CALCULATE STRESS-DISPLACEMENT TRANSFORM~TION MATRIX ( H'
e GIVEN BY C *
LNORT ,WHERE C IS THe STRESS-STRAIN ~ATRIX
C FOR THE RELEVANT MATERIAL TYP~ ( MAT'
C
DO 10C J=1,6
DO 100 1=1 ,3
H(I,J)= 0.0
DO 100 K=1,3
H(I,J)=H(I,J' + C(I,K,MAT,*LNORT(K,J'
1 CO CONTI NUE
C
IF (INTFR .EQ. 0) WRITE (6,998) «H(I,J),J=1,6',I=1,3)
999 FOQMAT (41HO ST~ESS-DISPLACE~ENT TRANSFORMATION MATRIX/
1 (5X,6E12.4',
C
C NOW PETURN IF THIS SUAROUTIN~ WAS CALLED FROM STRES
C
IF(KUT.EQ.1' RETURN
C
C ~OW COMPLETE THE FORMATION OF THE STIFFNESS MATRIX ( KSUB'
C FIRST ~ULTIPLY LNORT BY THE AREA OF THE ELE~ENT
C
AREA=AREA/~.O
00 200 J=1,6
DO 200 1=1,3
LNORT(1 ,J'=LNORT( I,J)*M~EA
2eo CONTINUE
C
C CALCULATE (LNORT) TRANSPOSE * H - EQUATION 24A IN TEXT
t
DO 300 J=1,6
DO 300 1=1,6
KSUB(I,J)=O.O
DO 300 K=1,3
KSU/3( I,J)=KSUB( I ,J'+LNOPT(K, I '*H(K,J)
300 CONTINUE
C
C CALCULATE CO~TRIBUTIONS OF GRAVITY LOADING, ACCELERATION LOADING
C AND WATER FORCES TO LOAD VECTOR -EQUATIONS 65 AND 66A
C
CONST=AREA/3.C
XCOMP=(RO(~AT'*ACELX-~PDX)*CONST
YCOMP=(RO(MAT)*ACELY-DPDY'*CONST
F( 1 )=XCOMP
F(2)=YCOMP
F(3)=XCOMP
F(4)=YCOMP
F (5)= XCOMP
F(6)=YCOMP
C
C NOW AD~ TH~ CONTRIAUTION OF THF RFSIDUAL STRESSES - - EONS. 27 AND 27A
C
IF(NRES.EQ.O' GO TO 500
DO 40C J=1,6
DO 40C 1=1,3
F(J)=FeJ)-LNORT(I,J)*RFSIO( I)
400 CONTINUE
C
5eo IF (INTER .EQ. 0) WRITE (6,997' M,«KSU8el,Jl,J=1,6),F(I),t=1,6'
997 FORMAT (29HO STIFFNESS MATRIX - ELEMENTIJ,73X,11HLOAD VECTOR/
1 (2X,6E12.4,28X,E12.4"
C
A ppendix 1 407

Rt'TU RN
E ND

SUAROUTINE J ST IF( MATT ,X,Y,RESIO)


C
C SUBPOUTINE TO CALCUL AT E S T IF F~ESS MAT~IX AND LOAD VECTOR rOR J O INT
C THIS SUBROU T INE IS DF- LIqERATELY L ONG WINOE n so THAT ALTERATIONS
C TO THE JOINT STIFFNESS MATRIX MAY BE MA DE WITH EASE
r
CO MM UN /STFF/ KJS( 23 ), K JN(2 :3 ) ,K J SNU ( 2 ] ) ,!(J NNU( 2 3) ,t TERl,I T E ~ N ,IT
RE AL KJ S ,KJN ,K J S NU, KJ NN U
CO MMON / CONS/ RO( 10) ,ACELX,ACELY,NRES,INTFR, C( 3, 3,IO )
C O ~ MON / F OP C / K5 U8( S, 8) ,F (S),LBA D ,M, H(3,6)
REA L KS UB
D I MENSIO N X(4 ) , Y(o\' ,RESIf)( 3 )
REAL L , KSO,K NO
C
C STAR T BY C HE CK IN G TH AT IT IS A VI\LI0 JOINT
C
IF ( X( 1) .IIIF. X ( 0\ , ) GO TO 10
IF ( X (2 ' .NE. X ( '3)) GO T I") 10
IF (Y ( 1 , .NF . Y ( " )) GO TO 1C
IF ( y ( 2) .NE. Y{ :3) , GO TO 10
GO TO 50
C
C ELEMENT DATA ERROR
C RETURN IF PREVIOUS ERRORS FOUND
C
1~ WRI T E (6, 2Q C~) M
2000 FORMAT ( 32HO JOINT FLEMENT DATA ELEMFNTI5)
LAAD= L BAD+l
PET URN
C
C SET MATT TO L OCAL VAR IABLE MAT (SPEEDS UP FXECU T ION'
C
50 MAT= MATT
C
( CALCUL ATE JOINT D IM F NSIONS L= JOINT LENGTf1
C
DX=X(2 ) -)« l '
OY='(2' -Y (I '
L=sa~T(DX*DX+ D Y*DY )

C
C CONS TANTS FOR FURTHFR CALCULATION
C
COSA=DX/L
SIN A=DY / L
S2= CO S4 *S INA
CC=COS. $ COSA
SS=SIN<\*SINA
C2=C C- SS
C
C ARE THERE RESIDUAL STRESSES - IF SO , TRANSFORM THEM
C
IF (NRES .GT. 0) GO TO 80
C
C RESI DUAL S HEAR S TR FSS I S ZERQ
C
TORSN=C.O
SIGMA = o.
408 Appendix 1

GO TO 90
80 TORSN=-S2*RESID(1)+S2*RESID(2t+C2*RESIO(3)
SIGMA= SS*RESID(l t+CC*RF.:SID(2)-2.0*S2*RESIDC3)
C
C
C RFTAIN TRANSFORMED STPESSES FOR FUTURE USE
C
90 RESID(1) = TORSN
RESID(2)=SIGMA
RFSIDC3'=0.0
C
IF (INTER .Ea. 0' WRITE C6,999) TORSN,SIGMA
999 FORMAT C31HO JOINT RESIDUAL SHEAR STRESS =E12.4/31H JOINT RESIDUA
1L ~ORMAL STRESS=E12.4,
C
C JOINT STIFFNF.:SS WILL HAVE BEEN READ FROM TAPEC7, IF THIS IS AN
C ACCELERATED RESTART. IF THE NORMAL RESTART WAS USED THIS SUBROUTIN=
C WILL NOT HAVE BEEN ENTERED
C
IFCITER1.GT.l) GO TO 95
C
C CALCULATE JOINT NORMAL STIFFNESS - EQUATION 44A
C

c
C NORMAL STIFFNESS MUST LIE WITHIN LIMITS 0.01 aU/VMC AND 100 QU/VMC
C
SLOMIN=-O.01*CC1,l,MAT)/CC2,2,MAT)
SLOMAX = 10000.*SLOMIN
IF (KJNCM'.GT.SLOMAX' KJNCM'=SLOMAX
IFCKJNCM'.LT.SLOMIN) KJN(M)= SLOMIN
C
C RECOVER JOINT SHEAR STIFFNFSS
C
KJSCM'=C(3,1,MAT)
c:
C ~3W SET UP JOINT STIFFNESS IN LOCAL COORDINATES - EQUATION 35
C
95 KSO = KJS(M'*L/4.0
KNO = KJN(M'*L/2.0
DO 100 J=1,8
DO 100 1=1,8
KSUBCI,Jt=O.O
100 CONTINUE
C
C NOTE THAT THE JOINT STIFFNESS MATRIX IS WRITTEN OUT IN FULL.
C THIS PERMITS READV ALTERATION TO A NON SVMMFTRIC MATRIX.
C SINCE A ST~AIGHT INVERSION OF THE GLOBAL STIFFNESS IS CARRIED OUT
C SUCH MODIFJCATIONS wOULD CAUSF NO PROBLEMS
C
KSUB C 1 , 1 , = KSO
KSUB(3,1)= KSO
KSUB(5,ll=-KSO
KSU9(7,1,=-KSO
KSUA(1,3'= KSO
KSURC3,3)= KSO
KSUBC5,3)=-KSO
KSUBC7,3,=-KSO
KSUBCl,5,=-KSO
KSUB(~,'5'=-KSO
KSUA(5,S'= KSO
KSUAC7,5'= KSO
KSUA(1,7'=-KSO
Appendix 1 409

KSUB(3,7)=-KSO
KSUA(5,7)= KSO
KSUB(7,7)= KSO
KSUB(2,2)= KNO
KSUB(2,8,=-KNO
KSUB(4,4,= KNO
KSUB(4,6,=-KNO
KSU[316,4)=-KNO
KSUA(6,6)= KNO
KSUB(8,2)=-KNO
KSUB(8,B)= KNO
C
IF (INTER .EO. 0) WRITF (6,998) «KSUB(I,J),J=I,8),I=I,a)
998 FO~MAT 137HO JOINT STIFFN~SS - LOCAL COO~OtNATES/(2X,8E12.4»
C
C
C PER~ORM MULTIPLICATION TO TRANSFORM FROM LOCAL TO GLOBAL COORDS
C ECONO~IES ACHIEVED OV AVOIDING ZEROS IN J7A
C USE OF SVMMETRV OF STIFFN~SS MATRIx IS AVOIDED HERE
C TRANSFORM IN 2*2 BLOCKS AT A TIME
C FOQ"1 (K,*(T) IN EQUATION 38
C
DO 200 1=1,4
IK=I+I
II=IK-l
('10 200 J=1,4
JK=J+J
JJ=JK-l
Tl1=KSUB(I I ,JJ)*COSA-KSUB( I I,JK)*SINA
T 12 =K SU B ( I I, J J ) * S I NA+KSUB I I I, JK ) *COSA
T21=KSUB(IK,JJ)*COSA-KSU[3(IK,JK)*SINA
T22=KSUB(IK,JJ)*SINA+KSUA(IK,JK)*COSA
C
C PR~MULTIPLY 8V (Tl TRANSPOSF EQUATION 38
C
KSUB( I I,JJ)=COSA*Tll-SINA*T21
KSU8(II,JK)=COSA*T12-SINA*T22
KSU8ITK,JJ)=SINA*Tl1+COSA*T21
KSUB(IK,JK)=SINA*T12+COS~*T22

200 CONTINUE
C
C ~OW CALCULATF LOA~S COR~ESPONDING TO RESIDUAL STRESSES
C CHANGE TO EFFECTIVE NORMAL STRESS AV ADDING WATER PRESSURE
C TRANSFORM TO GLOBAL COOROINATES IMMEDIATELV - EQUATION 40 COMPLETED
C
SIGMA=RESID(2)
L=L/2.0
Fxt =-L*TORSN*COSA + L*SIGMA*SINA
FYI =-L*TORSN*SINA - L*SIGMA*COSA
F(I)=-FXI
F(2,=-FVI
F(3,=-FXI
F(4)=-FYI
F(S)= FXl
F(6,= FVI
F(7)= FXI
F(B)= FVI
C
IF (INTER .EQ. 0) \II"HTE (6,997) M,(KSUA(I,J),J=I,B),FII),I=l,B'
997 FORMAT (61HO JOINT STIFFNESS AND FORCE VECTOR - GLOBAL CO OROS. - E
lLEMFNTI3/(2X,8EI2.4,EI4.4')
c
410 Appendix 1

RETURN
END

SUBROUTINE STRFS
c
C SUBROUTINE TO CALCULATE SOLID CLE~ENT STRESSES FRo~ DISPLACE~ENTS
C
COMMON /MFSH/ X(2S),Y(2S),R(SO),U(50),KS(SO,50),
1 NOD(2i,4"RESI0(23,3),SIGNO(23,3),~TYPEe23),
2 NODES,NELE~,NSOL,NFREE
REAL KS
COMMON /FORC/ KSUB(8,8),F(8),LBAO,M,H(3,6)
REAL KSUA
DIMENSION DISP(6),OUMMY(3),XT(4',YT(4),SIG(6)
c
C WRITE HEADING FOR PRINTOUT OF STRESSES
C
WRITEe6,20CO)
c
C CALCUL~TE STRESS IN EACH SOLID ELEMENT IN TURN
C
00 40C M=I,NELFM
C
C IS IT A SOLIO ELEMENT OR A JOINT
C
MAT=MTYPE(M)
IF (MAT .GT. NSOL) GO TO 400
("
C ASSEMBLE VECTOR OF NODAL CUORDINATES XT AND YT
C ALSO ASSEMBLE VECTOR (DISP) OF ELE~ENT NODAL D[SPLACE~ENTS
C
DO 250 [-:: 1, J
11=1+1-1
N= NOD ( ,14, I )
JJ=N+N-l
XT(I,=X(N'
YTn )=Y(N)
DISP(lI)=U(JJ)
DISP( 11+1 )=UeJJ+l)
250 CONTINUE
C
C NOW CALL SUBROUTINF TRIA TO FORM ELEMENT STRESS-DISPLACEMENT
C TRANSFORMATION MATRIX -(H)
C DUMMY IS USED TO FILL UP UNUSED PART OF THE PARAMETER LIST
C

C
C NOW CALCULATE STRESSES - ADDING INCREMENT TO RESIDUAL STR~SSES
C EQUATION 25 IN TEXT
C
DO 300 1=1,3
SIG( I )=RES[D(M,I)
DO 30e J=1,6
SIGC I )=SIG( [)+He I,J)*OISP(J)
300 CONTINUE
C
C CALCULATE PRINCIPAL STRESSES FROM XX(SIG(IJ),YY(SIG(2»,XY(SIGe3»
C
CC=(SIGC1)+SIG(2»/2.0
BB=(SIG(1)-SIGC21)/2.0
Appendix 1 411

CR=SQRTCAB*BA+SIG(3'*SIG(3»
SIG(4)=CC+CR
SIG(S)=CC-CR
SIG(6)=28.6q~*ATAN2(SIG(3,,8B)
C
C WRITE OUT ELEMENT STRESSES
C
WRrTE (6,2001) M,SIG
400 CONTINUE
c
RE TURN
C
2000 FOP.~AT (24HO SOLID ELEMENT STRESSES//tOX,7HELE~~NT6X,q~XX-STRESS7X
I,QHYY-STRESS7X,9HXY-STQESS6X,43~MAJOR PRINCIPAL MINOR PRINCIP~L Oq
2 IE I'IT A T I ON)
2001 FORMAT (It7,SEI6.4,FI4.2)
C
END

SUBROUTINE JSTR
C
C SUBROUTINE TO CALCULATE RESPONSE OF JOINTS TO NODAL DISPLACEMENTS.
C INCREMENTAL LOADS FOP NEXT ITERATION ARE AUTOMATICALLY CALCULATED.
C VECTOR R WILL FINALLY CONTAIN INCRE~ENTAL LOADS READY FOR
C CALCULATION OF NEW INCREMENTAL DISPLACEMENTS.
C
COM~ON /MESH/ X(2S),YC2S),R(SO),UC50),KSCSO,50),
1 NOD(23,4),RESID(23,3"SIGNO(23,3),~TYPE(23),
? NODES,NELEM,NSOL,NFREE
REAL KS
COM~ON /STFF/ KJS(23),KJN(23),KJSNU(23),KJNNU(23),ITER1,ITE~N,IT
REAL KJS,KJN,KJSNU,KJNNU
COMMON /CONS/ RO(10),ACELX,ACELy,NRES,INTER,C(3,3,1~)
DIMENSION LAB(4),U5N(B),CJ(9,lO),F(B)
EQUIVALENCE (CJ( 1,1) ,C( 1,1,1»
REAL L,LHLF,KSO,KNO,MO
C
WRITE(6,2000)
C
C CONVERSION CONSTANT FOR DEGREES TO RADIANS
C
CONV= ATAN(I.0)/45.0
C
C TAKE JOINTS ONE AT A TI~E

C
DO 800 M=l,NELEM
C
C SELECT JOINT TYPE
C
MAT=~TYPECM)

c
C IS IT A JOINT - IF NOT GO TO NEXT ELEMENT
C
IF (MAT .LE. NSOL) GO TO 800
C
C RECOVER BASIC JOINT INFORMATION READY FOR CALCULATION
C
I I =NOD (M, t )
JJ=NOO(M,2)
412 Appendix 1

DX=X(JJ)-X( II)
DY='t'(JJ)-'t'(II)
C
C JOINT LENGTH AND ORIENTATION
C
L=SORT(DX*DX+O't'*D't"
COSA=DX/L
SINA=DY/L
C
C POSITION VECT3R USFD TO IN TRANFER OF DATA TO AND FRO~ GLOBAL VECTORS
C
no 200 1=1,4
J=NOD (M, I )
JJ=J+J-l
LAB([)=JJ
11=1+1
C
C TRANSFORM OISPLACEMENTS TO LOCAL COORDINATES
C
USN(I[-l'= U(JJ)*COSA+U(JJ+t )*SINA
USN(II) =-U(JJI*SINA+U(JJ+l)*COSA
2CC CUNTINUE
C
IF (INTER .EO. 0) WRITE (6,999' USN
999 FORMAT (47HO JOINT OISPLACF~ENT VECTOR - LOCAL COOROINATFS/
1 (2X,8F12.4»
C
C
C JOINT MAXIMUM CLOSURE - FOUATION 44
C
VM = -CJ(S,MAT)
IF(RESID(M,2).LT.O.1 VM = -CJ(S,MAT).CJ(6,MAT)/RES{O(M,2'
C
C JOINT SHEAR AND NORMAL STIFFNESSES
C
KNO KJN(M)
KS[) KJS(M)
C
C RATIO OF RESIDUAL TO PEAK SHFAR STRNGTH
C
80 = CJ(4,MAT)
C
C COMPPESSIVE STRENGTH OF WALL ROCK
C
au = CJ(l,MAT)
C
C RATIO OF COMPRESSIVE TO TENSILE STRENGTH LADANYI ANO ARCH TERM N
C
RAT 1/CJ(2,MATI
C
C TANGENT OF ANGLE OF FRICTION OF A SMOOTH JOINT
C
TANTH = TAN(CJ(7,MAT)*CONV)
C
C TANGFNT OF DILATATION ANGLE - ZERO CONFINING PRESSURE
C
TANI = TAN(CJ(B,MAT)*CONV)
C
C CALCULATE THE JOINT DEFORMATION - TOTAL DISPLACEMENTS - EQUATION 28,29
C
UO (-USN(1)-USN(3,+USN(S'+USN(7,)/2.0
VO (-USN(2)-USN(4)+USN(6)+USN(8')/2.0
W (USN(2)-USN(4)+USN(6)-USN(8»/L
Appendix 1 413

C
C CALCULATE TOTAL STRESS THIS ITERATION - EQUATION 3211.,25,33
C TNS=SHEAR STRESS, SN=NORMAL STRESS, MO=MOM~NT
C
TNS Kso*un +SIGNO(M,I)
SN KNO*VO +SIGNO(M,2'
MO L*L*L*KNn*W/~.O+SIGNU(M,3'
C
WPITE(6,2001IM,UD,VO,W,TNS,SN,MO
c
C ~DW TAKE EACH END OF THE JOINT IN TURN AND CH~CK THE SHEAR AND
C NORMAL FORCES FOR EACH NODAL PAIR. FIRST I,L NODES ,THEN J,K
C VAqlABLE SIGNE IS USED TO CONTROL THE SIGN OF TERMS IN THE B MATRIX
C (EQUATION 31' THIS MA TRIX IS NOT SFT UP BUT THE RELATIONSHIPS IT
C DEFINES ARE USED
C FIRST SET L TO HALF LENGTH OF JOINT
C
LHLF = L / 2.C
SIGNE ::: 1.0
TORM =0.0
FNOM O.C
FNIM 0.0
C
00 550 I::: 1,2
SIG NE ::: -SIG NE
c
C CALCULATE NODAL FORCES FOR FIRST,PREVIOUS AND CURRENT ITERATIONS
C (EQ 31 IN TEXT) THESE ARE EXTERNAL FORCES AT L AND THEN AT K
C
C
C F~IL NOR"'IAL FORCE FOR LAST ITERI\TION
C FSO INITIAL SHEAR FORCF
C FNO INITIAL NORMAL FORCF
C FSI CURRENT SHEAR FORCE
C FNI CURRE'NT NORMAL FORCE
C
FNIL SIGNO(M,21*LHLF + SIGNE*SIGNO(M,3)/L
FNO RESID(M,2)*LHLF + SIGNE*RESIO(M,3'/L
FSD RES I D ( M, 1 I * L HL F
FSI TNS * LHLF
FNI SN*LHLF + SIGNE*MO/L
C
C NORMAL FORCE PEq UNIT AqEA FOR THIS HALF OF THE JOINT IS SIGMA
C
SIGMI\ == FNI/LHLF
c
C SHEAR STRENGTH ETC. SET TO ZERO IF JOINT IN TENSION
C
IF(SIGMA.LT.O.OI GO TO 305
TORP==C.O
TORR=O.O
TOR =0.0
UP =0.0
UR =C.C
DILAT=J.O
GO TO ~60
C
C CA.CULATE PEAK SHEAR STRENGTH - EQUATIONS I~ TO 16 OF CHI\PTER 5
C F IRST CHECK IF NDR"'IAL LOADING IS ABOVE THE COMPRESSIVE STRFNGTH
C
305 IF (au .LT. SIGMA' GO TO 320
C
C TRANSITION STQESS EXCEEDED
414 Append~ 1

C
AS=l.O
VDOT=O.O
C
C LIMIT SIGMA/au RATIO TO 1.0 SO THAT PEAK SHEAR STRENGTH IS LI~ITED
C
RATIO = 1.0
GO TO 330
C
C BE_OW TRANSITION STRESS - EaUATIONS 16A AND 16~ OF CHAPTER 5
C
320 RATIO=SIG~A/aU
AS=1.C-SQRT«l.C-RATIO)**3'
VDOT=(1.0-RATIO'**4*TANI
C
C EVALUAT~ EQUATION 14 OF CHAPTER 5 TORP IS PEAK SHEAR STRENGTH
C
330 TERMl SIGMA*(l.O-AS)*(VnOT+TANTH)
TER~2 AS*QU*(SQRT(RAT+l.O)-1.0'/RAT*saRT(1.O+RAT*RATIO)
TERM3 1.O-(1.0-AS,*VDOT*TANTH
TORP=AAS(TER~1+TERM2'/TER~3
C
C IF TRANSITION STRESS EXCEEDED THEN NO PEAK-RESIDUAL BEHAVIOUR
C TORR = RESIDUAL SHEAR STRENGTH FROM EQUATION 60
C
TORR = TORP
IF (QU .LT. SIGMA) TORR = TORP*(Bn + (I.C-80)*SIG~A/aU'
C
C SE~ECT PEAK AND RESIDUAL SHEAR DISPLACEMENT - EQUATION 54 AND 54 A
C SIGN OF DISPLACE~ENT DETERMINES THE EQU4TION- SET SEL TO THE SIGN
C
AUO AAS(UO)
SEL UO/AUO
C
C UP PEAK SHEAR DISPLACEMENT
C UR RESIDUAL SHEAR OISPLACE~ENT
C
UP (SEL*TORP-RESID(M,l)}/KSO
UP (SEL*4.0*TORP-RESID(M,1»)/KSO
C
C SELECT RANGE TO WHICH SHEAR DISPLACE~ENT BELONGS
C AND CALCULATE LIMITING SHEAR STRESS CORRESPONDING TO DISPLACEMENT EQS5
C
IF (AUO .GT. ABS(UP') GO TO 400
C
C IN LINEAR R~NGE DFFINED-BY SHFAR STIFFNESS - RANGE III
C
TOR = KSO*UO + RESID(M,I)
GO TO 450
C
400 IF (AUO .GE. ABS(UR)' GO TO 430
C
C IN FALLING PORTION - RANGE II OR IV
C
TOR = SEL*(TOPP-(UO-UP}/(UP-UP'*(TORP-TORR)'
GO TO 450
C
C IN RESIDUAL PORTION - RANGE I OR V
C
430 TOR SEL*TORR
AUU ABS(UR)
C
C NOW CALCULATE DILATION - FIRST SET AUO TO ABS(UR' IF SHEAR
Appendix 1 415

C BFHAVIOUR NON - L INe AR - SEE EQU ATION 62 (WIL L BF. AAS ( UO ) IF NOT ~ESE T
C
450 DI L AT (F Nf L/ (L HLF *Q U ' -1.OI **4 *TA N I *( AU O+A 8S (F SO )/(KS O*LHLF »)
C
C CAL CULATE SHEAR FO RC E TO RE APPLIED NEXT ITERATI ON -EQU AT I ON 57
C
46 0 OELS TOR*LHLF - FS I
C
C CALCULATE WHE T HER NODAL PA IR OPE N ING OR CL OS IN G
C SET DELV TO TOT AL NORM AL DISPLACE~E NT - E aUATI O~ 4 7 A
c
I1= Y+l
IJ= 10- 11
DEL V= - USN( It '+ U5 N( [J)
I F ( DEL V) 520,5 10 ,500
C
C JOINT OPENING - EQUATION 47A
C
5eo DEL N = CDEL V/ (V M- DE LV ) + 1 .0 , *FN O - F~ f
GO TO 530
C
C JOINT UNCH AN GED
C
510 DE LN = 0.0
GO TO 530
C
C JO IN T CLOSI NG - EQUATION 50A
C
5 20 DE LN (DELV -V M* <FNI- F NO)/FNI ' *KNO*LHLF
C
C C O~PU TF EXTER NAL FORCES ON JOINT Ea UIV~ LEN T TO INCR E ASE OF RE SIDU_L
C NO RMAL STRESS - CA USED 3V DILA T I ON
C
530 DILN - b l~ A T* F NI/(V~)
C
C NOW TRA NSF ER NO DA L F ORCES TO J OI NT FO RC E VECTOR
C
F (I I) = -OEL N-OILN
F( JI -1 ) = -DELS
Fe I JI = DELN+-D ILN
FCIJ-l):;:DF.LS
C
IF (INTFR .E O.~) WR IT E (6,998) lIP, UR,TORP,TOR R ,TOR ,I) ILA T, DE LV
99B rOR MA T (5HO UP=E 1 2. 4 ,4H UQ = E1 2. 4,6H TORP =E 12 .4, 6H TORR = E1 2.4.
1 5~ TO R=E 12.4 / 8H DILA T= E12. 4 ,6H DELV=E12.4)
C
C THE F OLL OWING ARE U SED TO S AV E MEAN SHEAR STRESSFS AN D NOR MAL Fa RCE S
C FO R CALCUL~TION OF NE~ JO I NT STIFFNESSES
C
TO RM TO RM. T OR/2.0
F NIM FN IM + F~I
F NUM FNOM +- FNO
S S O CO NTI NUE
C
IF ( IN TER .EO. 0) WRITE ( 6 t 9CJ7) F
997 FOR MA T (30HC INCREME NT AL FORCES F OR J OI NT/ l X, 8F. 12 .4 )
c
C
C CA L CULATE I NI TIAL S TR E SSE S CORRESPONDI NG TO P REVIOUS PLUS
C I NCREMENT J UST C ALCULATEO ( EQUATIONS 30 I
C L IS BA CK TO FULL JOINT LENGTH
C
S IGN O ( M,l ) SIGNO( M,U +-(F( 5)+ F(7 1. /L
416 Appendix 1

SIGNO(M,2' SIGNO(M,2) +(F(6)+F(B)'/L


SlGN(J(M,3' SIGNO(M,3) +(F(6,-F(B')*LHLF
C
C ROT4TE rORCES TO GLOBAL COORDINATE 4ND ADO TO GLOBAL FORCE V~CTOR R
C - SEE EQUATION 37
C
DO 60e 1=1,4
11=1+1-1
JJ=LAB(I'
R(JJ) = R(JJ' - F(II)*(OSA + F( 11+1 '*SIN4
R(JJ+l)=R(JJ+l '-F(I n*SJNA - F( [[+1 t*COSA
600 CONTINUE
c
IF(IT.LT.ITERN' GO TO BOO
c
C SET NEW JOINT SHEAR STIFFNESS
C
KJSNU(M) = (TORM-RESID(M,t')/UO
IF(DELV' 700,BOO,750
c
C JOINT CLOSING - NEW JOINT NORMAL STIFFNESS
C
700 KJNNU(M, = FNIM/(VM*L)
GO TO BOO
c
C JOINT OPENING - SET NEW JOINT NORMAL STIFFNESS
C
750 KJNNU(M) = FNOM/«VM-DELV'*L)
C
8CO CONTINUE
C
IF (INTER .NE. 0' GO TO 900
WR I T E ( 6 , 994 ) (K J S ( I ) ,K J N ( I , , I =1 ,N E L EM'
994 FORMAT (19HO JOINT STIFFNESSES/(2X,IOEI2.4,)
c
WRITE(6,996) (SIGNO(I,I),SIGND(I,2),SIGNO(I,3),I=I,NELEM'
996 FORMAT (38HO INITIAL STRESSES AT END OF INCREMENT/
1 2(6X,5HSHEAR6X,6HNORMAL6X,6HMOMENTllX'/
2 (2X,3EI2.4,IOX,3EI2.4"
C
WRITE (6,995) (R(I), 1=I,NFREE)
995 FORMAT (25H0 INCREMENTAL LOAD VECTOR/(2X,IOEI2.4),
C
900 RETURN
C
20CO FORMAT (IH025X,19HJOINT DISPLACEMENTSI6X,14HJOINT STRESSES/7X,
17HELEMENT5X,5HSHEAR6X,6HNORMAL5X,BHROTATION6X,5HSHEAR6X,6HNORMAL6X
2,6HMOMENT'
2001 FORMAT (ll0,2X,6EI2.4)
C
END
. appendix

conversion factors

Abbreviations are shown in parentheses; M


I.


L~ ...
I ~

! LENGTH
1 inch (in) 25.4 millimeters (rnrn)
f., 1 foot (ft) 0.3048 meters (m)
1 mile 1.60934 kilometers (Km)

AREA
2 4 2
1 in 6.4516 x 10- m
2
1 ft2 0.092903 m
2
1 acre 4,046.86 m
I VOLUME
t' 1 fluid ounce (U. S. ) 2.95735 x 10- 5 m
3

1 liter 1 x 10- 3 m3
3 3
1 U.S. gallon 3.7854 x 10- m
3 3
1 U.K. gallon 4.5461 x 10- m

FLOW RATE
3
1 ft 3 /sec 0.028317 m /sec
-5 3
1 U.S. gallon/min 6.30902 x 10 m /sec
418 Appendix 2

FORCE
1 dyne 1 x 10- 5 newtons (N)
1 ounce 0.278014 N
1 pound (lb) 4.448222 N
1 kilogram (Kg) 9.80665 N
1 U.S. ton 8.89644 KN

PRESSURE
2 47.88026 N/m2
1 pound/foot (psf)
1 foot of water (at
60 degrees F) 2.9861 KN/m 2
1 pound/inch 2 (psi) 6.89476 KN/m2
2
1 Kg/cm 0.0980665 MN/m2
2
1 bar 0.1 MN/m

Note: 1 MN/m2 1 MPa (MegaPascal) = 145.037 psi


2
For rough calculations, 1 bar = 1 Kg/cm 2 = 1 ton/ft
re e.r e ces

A list of a bbreviations used follows the li st of re fe r e nces. The


numbers to the left of each entry identify the chapters to which the
article is most pertinent. General references are assign ed n umber 1.
Government documents (u.s.) are available from NTIS, Sprin g f ield, Va.
22151; the AD number identifies the documen t for NTIS.

2 Aastrop, A., and Sallstrom, S. (1964), "Further treatment of


problematic rock foundation at Bergeforsen Dam" Proc . 8t h Cong o on
Large Dams, Edinburgh, p 627

8 Agarwal, R.K., and Boshkov, S.H. ( 196 9), "Stresses and d isplac e-
ments around a circular tunnel in a three layer medium" I JRM&M S, V6 n 6

3 Arnold, K.J. (1941), "On spherical probability d is t ribut i ons"


PhD Thesis, M.I.T.

5 Ashby, J. (1971), "Sliding and toppling modes of fa il u re in


models and jointed rock slopes" MsC Thesis, Imperial College , London

3,4 Badgeley, P.C. (1959), "Structural problems for the exploration


geologist II (Harper)

8 Baker, L.E., Sandhu, R.S., and Shieh, W.Y. (1969), "Appli c a tion
of elasto-plastic analysis in rock mechanics by the finite element
method, Proc. 11th Symp. on Rock Mech . , p237
6 Banks, D.C., and Strohm, W.E. (1974), "Calculations of r o ck
slide velocities" Proc. 3rd Congo ISRM, Denver, V2B, p839

4 Barr, D.J. (1969) "Use of side looking air -bo rne ra dar (SLAR)
imagery for engineering soils studies", U.S. Army Eng. Topog. Lab . ,
Ft. Belvoir, Va., Tech. Report 46-TR (AD 701-902)
Goodman-G eol ogical Eng. - 28
420 References

3 Barton, C.M. (1974), "Simplified procedures for the vector sum-


mation and statistical analysis of spherically distributed point
clusters" CSIRO (Australia) Div. of Appl. Geomechanics, Tech. Rep. n20

7 Barton, N.R. (1971a), "A model study of the behaviour of steep


exc-avated rock slopes" PhD thesis, Imperial College, London

5 Barton, N.R. (1971b), "A relationship between joint roughness


and joint shear strength" Proc. Int. Symp. on Rock Fracture, Nancy,
(ISRM), paper 1-8

5 Barton, N.R. (1972), "A model study of rock-joint deformation"


IJRM&MS V9 n5

5 Barton, N.R. (1974a), "A review of the shear strength of filled


discontinuities in rock" Saertrykk, Fjellsprengningsteknikk Bergmeka-
nikk, 1973,(TAPIR, Norway) - Broch, Heltzen, and Johannesen, ed.
Chapter 19

5 Barton, N.R. (1974b), "Estimating the shear strength of rock


jOints" Proc. 3rd Congo ISRM, Denver, V2A, p219

2 Barton, N., Lien, R., and Lunde, J. (1975), "Analysis of rock


quality and support practise in tunneling and a guide for estimating
support requirements" Rock Mech. (in press)

8 Baudendistel, M. (1972), "Interaction between the tunnel lining


and the surrounding rock" (in German', Veroeffent lichungen Inst. Bod-
enmechanik und Felsmechanik, Univ. Fredericana, Karlsruhe

8 Baudendistel, M., Malina, H" and Muller, L. (1970), "The effect


of the geologic structure on the stability of an underground power
house" (in German), Proc. 2nd Cong. ISRM, Belgrade, V2, paper 4-56

1 Bellier, J. (1967), "Le Barrage de Malpasset", Travaux, July

8 Benson, R.P., Sigvaldason, O.T., and Kierans, T.W. (1970),


"In-situ and induced stresses at the Churchill Falls underground power
house, Labrador" Proc. 2nd Cong. ISRM, Belgrade, V2, paper 4-60

4 Berents, H. P. (1961), "A retractable triple tube core barrel II


Snowy Mtn Hydroelectric Authority, Cooma, Australia

2 Bergh-Christensen, J., and Selmer-Olsen, R. (1970), "On the


resistance to blasting in tunnels" Proc. 2nd Congo ISRM, Belgrade,
V3, paper 5-7

1,2 Bernaix, J. (1966), "Contribution a l'etude de la stabilite des


appuis de barrages, Etude geotechnique de la roche de Malpasset" PhD
Thesis, Ecole Poly technique , Paris (Dunod) -

2 Bernaix, J. (1969), "New laboratory methods of studying the


mechanical properties of rocks" IJRM&MS V6, p43
References 42 1

5 Bernaix, J. (1974), "Properties of rock and rock masses" Proc.


3rd Congo ISRM, Denver, V1A, p9

8 Best, B.S. (1970), "An investigation into the use of finite


element methods for analysing stress distributions in block jointed
rock masses" PhD Thesis, James Cook Univ. of N. Queensland, Towns-
ville, Australia

8 Bhattacharya, K., and Boshkov, S.H. (1970), "Determination of


the stresses and displacements in slopes by the finite element method"
Proc. 2nd Congo ISRM , Belgrade, V3, paper 7- 1 0

2 Bieniawski, Z.T. (1974a), "Geomechanics classification of rock


masses and its application in tunneling" Proc. 3rd Congo ISRM, Denver
V2A, p27

2 Bieniawski, Z.T. (1974b), "Engineering classification of jointed


rock masses, Auth or's reply to discussion", Trans. S.Afr. Inst. of
Civil En g ., July

5 Bishop, A.W. (1966), "Soils and soft rocks as en g ineering ma-


terials" Innaugura l Lecture, Imperial College, London

5 Bishop, A. W. (1973), " The influence of an undrained change in


stress on the pore pressure in porous media of low compressibility"
Geotechnique, V23, p435

2 Bjerrum, L. (1967) "Mechanism of progressive failure in slopes


of over-consolidated plastic clay and clay shales" the Third Ter z a ghi
Lecture, J. SM&FD, ASCE V93, p3

2 Bj e rrum, L., Br ekke, T.L., Mourn, J., and Selmer-Olsen, R. (1963),


"Some Norwegian studies and experiences with swelling materials in
rock gouges" Rock Mech. and Eng. Geol. V1, p23

8 Blake, W. (1966), "Applic. of the finite element method of an-


alysis in solving boundary value problems in rock mechanics" IJRM&MS,
V3, p169

8 Blake, W. (1968), "Finite element model study of slope modifica-


tion at the Kimbley pit", Trans. SME of AIME, B241, p525

8 Blake, W. (1971), "Rockburst research at the Galena mine,


Wallace, Idaho U.S . B.Mines Tech. Prog. Rep. 39

4 Blyth, F . G.H. (1965), "Geologic maps and their interpretation"


(Edward Arnold)

4 Bolstad, D.D. and Mahtab, M.A. (1974), "A Bureau of Mines direct
reading azimuth protractor", U.S.B. Mines Inf . Circ. 8617

5 Bowden, F. P., and Tabor, D. (1965), "The friction and lubrica-


tion of solids" (Clarendon Press)
422 References

2 Bradley, D.E. (1954), "Replica techniques" Brit. J. of Appl.


Physics, V5, p165

5 Brawner, C. 0., Pentz, D. L., and Sharp, J. C. (1972) , "Stability


studies of a footwall slope in layered coal deposit" Proc. 13th Symp.
on Rock Mech. (ASCE) p329

5 Bray, J.W. (1967), "A study of jointed and fractured rock" Rock
Mech. and Eng. Geol., V5, n2 and n3

2,5 Brekke, T.L. and Howard, T.R. (1972), "Stability problems caused
by seams and faults" Proc. 1st Rapid Excav. and Tunneling Conf. (AIME)
V1, p25

2,5 Brekke, T.L., and Howard, T.R. (1973), "Functional classifica-


tion of gouge material from seams and faults as related to stability
problems in underground openings" Report from Univ. of Cal. to U.S.
B. Mines, (AD-766-046)

2 Broch, E., and Franklin, J .A. (1972), "The point load strength
test" IJRM&MS, V9, p669

5,7 Brown, E.T. (1968), "The influence bf planar discontinuities on


the shear strength of a rock-like material" PhD thesis, James Cook
University of N. Queensland, Townsville, Australia

5,7 Brown, E.T. (1970a), "Modes of failure in jointed rock masses"


Proc. 2nd Congo ISRM, Belgrade, V2, paper 3-42
5 Brown, E. T. (1970b), IIStrength of models of rock with intermi t-
tent joints I,' J. SM&FD, ASCE, V96, p1935

8 Brown, R.E. (1968), "A multi-layered finite element model for


predicting mine subsidence ll PhD thesis, Carnegie Mellon Univ.,
Pittsburgh, Pa.

2 Brune, G. (1965), IIAnhydri te and gypsum problems in engineering


geology" Engineering Geology, Bull. A.E.G., V2, n1

3 Bucher, W.H. (1944), lIThe stereographic projection, a handy tool


for the practical geologist" Jour. Geol. V52, p191

8 Burman, B.C. (1971), "A numerical approach to the mechanics of


discontinua", PhD Thesis, James Cook Univ. of N. Queensland, Towns-
ville, Australia

2 Burton, A. N. (1965), "Discussion to a paper by D . Coates II,


IJRM&MS, V1, p105

5 Byerlee, J.D. (1967a), "Frictional characteristics of granite


under high confining pressure II , J. Geophysical Res. V 72, p3639

5 Byerlee, J.D. (1967b), "Theory of friction based on brittle


fracture" J. Applied Physics, V38, p2928
References 423

8 Byrn e, R .J . ( 1974L "Phy s ical and nume rical mo dels in rock


a nd s oil s lope s tab i l it y" PhD Thesi s, James Cook Vniv. of N. Queen s-
l and, Townsvil le , Aus t r a lia

8 Byskov , E. ( 1970) , "Th e calculati on of stre ss i ntens ity f actors


us in g t h e fi n it e e leme nt met hod- with cracked eleme nts" ~J. of
Fracture Mech. V6, p159

8 Call , R.O. (1971) , " Slope st ability study of the Tazadit pit,
Mauritan i a " Proc. 2nd Int. Coni. on Stability f or Open P i t Mines,
Vancouve r ( AIME)

8 Carnahan , B. , Luthe r, H., a nd Wi lke s, J. (1969) "App lie d


nume r i ca l met hods" ( John Wiley )

4 Cas s ines > R . ( 1 9 72) , "Re mo te s e nsing - - an e va l u a t ion of its


impa ct on e ar t h s ci ences ll Ge ophysical Prosp e ct i ng , V2 0, p142

2,4 Caterp i l l a r Tracto r Co. (1966) , "Handbook of ripp i ng , a g uide


to grea ter profi t s ll Third edition (Peoria, Il lin o is»

8 Chang! C. Y. a nd Nair, K. (1974) "Deve lopmen t a nd appl i cation of


a genera l compute r program fo r evaluating st abili t y of op eni ngs in
ro ck Tl Proc. 3rd Congo ISRM, De nver, V2B , p981

5,7 Chappel , B.A. (1972) , uThe mechanics of blocky mat eri al" PhD
thesis , Australian Nationa l Vniv ., Canberra

7,8 Chappel, B.A . ( 1974) ,. "Numeri c a l and p hy s i ca l e xpe riment s with


discon t inua" Proc. 3 rd Congo ISRM, Denver , V2A, p118

5 Chenevert, M.E. and Ga t li n,. C. (1965) , " Mechani cal aniso tropies
of l a minat e d s edi me n tary rocks '" SPE J o urn al ( AIME ) , V5, p67
8 Chris t iansen, L.M. , Mis te r e k, D .L., and Bowl e s, G.F . ( 1971 ),
"Foundat i on analysis of Auburn da m s ite" F roc. Int. Symp. on Ro c k
Fracture, Nancy (ISRM) pap er 2-26

8 Clough , R.W. (1960), "The f inite elemen t me th od in plane st r es s


analy sis ", Proc. 2nd ASCE Conf. on Electronic Comput a tio n, p345

8 Clough, R.W. (1965) , "Th e finite element me t hod in structura l


analysis " , Chapt e r 7 in St re s s Analysis, Zienkiewicz a nd Ho lister, eds.

4 Cluff , L.S . , and Slemmo ns , D.B . (1971) , "Was atch f au lt zon e


f e atures de f ined by low sun angl e pho t ography ~1 in Environme n tal
Geology of the Wasatch Fron t , Ut ah Geol. Assoc. PUb . n1 (U t ah Geo l.
and Min e ralog i cal Survey) pGl

2 Coates , D.F . ( 1964) , If Classification o f rocks fo r ro ck mechanics "


IJRM&MS, VI, p 421

1 Coates , D , F. (1967), "Rock me chanics pr in ci ples I I Ca nadian


Depart ment of Energy, Mines, and Resources, Monograph 874
424 References

2 Coates, D.F. and Parsons, R.C. (1966), "Experimental criteria


for classification of rock substances" IJRM&MS V3, p181

4 Colvocoresses, A.P. (1970), "ERTS - a satellite imagery"


Photogrammetric Engineering, V36, p555

2 Cording, E. J. and Mahar, J. W. (1974), "The effect of natural


geological discontinuities on behavior of rock in tunnels" Proc.
2nd Rapid Excav. and Tunneling Conf. (AIME) V1, p107

4 Cornea, I, and Enescu, D. (1965), "Seismic prospecting of faults


by means of apparatus located on a radial profile" (in French), Revue
Roumaine Geol., Geog., and Geoph, V4, p191

5 Coulson, J.H. (1972), "Shear strength of flat surfaces in rock"


Proc. 13th Symp. on Rock Mech. (ASCE) p77

4 Cumming, J.D. (1956), "Diamond drill handbook" (J.K. Smit,


Toronto)

8 Cundall, P. (1971), "A computer model for simulating progressive


large scale movements in blocky rock systems" Proc. Int. Symp. on Rock
Fracture, Nancy (ISRM), paper 2-8

8 Cundall, P.A. (1974), "A computer model for rock mass behavior
using interactive graphics", U.S. Army, Corps of Engineers, Technical
Report MRD 2-74 (Missouri River Division)

7 Currie, J. B. (1966), "Experimental structural geology", Earth


Science Reviews, V1, p51

8 Dahl, H.D. (1972), "Two and three dimensional elastic and elasto-
plastic analyses" Proc. 5th Int. Strata Control Conf., London

8 Dahl, H. D., and Voight, B. (1969), "Isotropic and anisotropic


plastic yield associated with cylindrical underground excavations",
Proc. lot. Symp. on Large Permanent Unde"rground Openings, Oslo (ISRM)

8 Dahl, H. D., and Parsons, R. C. (1972), "Ground control studies in


the Humphrey no. 7 mine", Trans. Soc.Min. Eng of AIME, V252, p211

2,4 Dearman, W.R., and Fookes, P. (1972), "Engineering geological


mapping in civil engineering practise" Qtly J. of Eng. Geol, V5, n4

4 De. Chambrier, P. (1953), "The microlog continuous dipmeter",


Geophysics, V18, p929

2,4 Deere, D.U. (1963), Tech. description of rock cores for eng.
purposes", Rock Mech. and Eng. Geol., V1, p18

2,4 Deere, D.U. (1968), "Geological considerations" Chapter 1 in


Rock Mech. in Eng. Practise, Stagg and Zienkiewicz, eds. (John Wiley)
References 425

1 Deere, D.U., Hendron , A.J., Patton, F.D., and Co rd i ng , E. J . ,


(1967), "Design of surface and n e a r sur£ ace const ru c t ion i n rock ",
Proc. 8th Symp. on Rock Mech. (AIME ), p 2 37

2,5 Deere, D.U., and Miller, R.P. (196 6), "Eng. cl as s ification and
index properties for intact rock" Ai r Force Weap o ns La b Tech Rep . AF
WL-TR-65-116

2 Deere, D.U., a nd Patton, F.D. ( 1971 ), "S l o p e s t a b ility in


residual soils", Pro c . 4th Pan Ame rican Conf . on So i l Mechan i cs and
Foundation Engineer ing, San Juan, p87

8 Dejean, M., Mart in , F., and Ra f fou x, J .F. (~97 0 ) , "Deformat i ons
of rock galleries during driving" (i n Fren ch ), P r o c. 2nd Congo ISRM,
Belgrade, V2, paper 4-36

8 de la Cruz, R. V., and Goodman, R. E. (1 969), "The bore h ole de e p -


ening method of stress measurement", Pro c . I nt . Symp . on De t . of
Stresses in Rock Masses , Lisbon (ISRM) p. 230

4 Denny, C.S., Warren, C.R., Dow, D.H., a nd Da le , W.J. (1968),


"A descr i ptive catalog o f sele c ted a er ial photog r aph s o f g eo logi c
features in the United States", U . S . G . S . P r o f. Pap er 590

8 deRouvra y , A.L., and Goodman , R.E. ( 1972) , "Finit e e leme nt


analysis of crack ini tia t ion in a b lo ck model e x pe riment " Rock Mech. ,
V4, p203

8 Desai, C.S., and Reese, L.C. (19 70), ' rS tr e ss- d e f orm a t ion a nd
stabili ty analyses of deep boreholes", Pr oc . 2 nd Co ng . ISRM , (Belgra d e ,
V2, paper 4-13

8 Desai, C. S., and Abel, J.F. ( 1972 ), "Intra. to the finite element
method" (Van Nostrand Reinhold )

4 Dobrin, M.B. (196 0 ), "Intro. to g eo p h y sic a l prospecting , 2nd


edition II (McGraw Hill)

8 Dolcetta, M. (1971), "Problems wi t h large underground s t a t i o n s


in Italy", Proc. Symp. on Underg roun d Rock Ch a mbers, (ASGE), p243

8 Dolcetta, M. (19 72 ), "Rock l oa d o n the sup p ort struc t ur e s of two


large underground hydroelectr i c power pl an t s ", P r o c. I n t. Symp on
Underground Op enings, Luce r n e ( I SRM)

5 Donath, F.A. (1964) , "Stre n gtlt var i a tion an d de formational


behavior in a n isotropic rock" in St a t e of Stre ss in t he Ear t h's Crust ,
W.R. Judd , ed. (Elsevier), p2 8 1

5 Donath, F. A. (1968), "Ex perime ntal s t udy o f ki n k-band deve lopme n t


in Ma r tinsb u rg s lat e ", Pro c . Con f. on Rese arch i n Te c t on i cs, Geo l .
Survey of Canada, Paper 68-52, p 25 5 ,
426 References

3,4 Donn, W.L., and Shimer, J.A. (1958), "Graphic methods in struc-
tural geology" (Appleton-Century-Crofts)

8 Dorn, W. S., and McCracken, D.D. (1972), "Numerical methods with


Fortran IV case st udies" (Wi ley)

2 Douglas, P., and Voight, B. (1969), "Anisotropy of granites: a


reflection of microscopic fabric", Geotechnique, V19, p376

8 Dubois, J. L. (1972), "The hyperplane perturbation method for the


analysis of non-linearities", PhD Thesis, Univ. of California, Berk.

2 Duncan, N. (1969), "Engineering geology and rock mechanics",


(Hill Books, Int. Textbook Cg., London, S.W. 1) 2 volumes.

8 Duncan, J.M., and Chang, C.Y. (1970), "Non-linear analysis of


stress and strain in soils", J.SM&FD,ASCE, V96, nSM5

4 Eaton, G.P., Martin, N.W., and Murphy, M.A. (1964), "Applic. of


gravity measurements to some problems in engineering geology", Engin.
Geology, Bull. A.E.G., V1, p6

2 Ege, J .R. (1968), "Stability index for underground structures


in granitic rock", Geol. Soc. America Memoir 110

7 Einstein, H.H., Nelson, R.A., Bruhn, R.W., and Hirschfeld, R.C.,


(1970), :Model studies of jointed rock behavior", Proc. 11th Symp. on
Rock Mech. (AIME), p83

5 Engelder, J. T. (1974), "Coefficients of friction for sandstone


sliding on quartz gouge", Proc. 3rd Congo ISRM, Denver, V2A, p499

8 Ergatoudis, I, Irons, B.M., and Zienkiewicz, O.C., (1968),"Curved


isoparametric quadrilateral elements for finite element analysis",
Int. J. of Solids and Structures, V4, n1, p31

7 Erguvanli, K.A., and Goodman, R.E. (1972), "Applic. of models to


engineering geology for rock excavations", Bull. A.E.G., V9, p89

5 Eurenius, J., and Fagerstrom, H. (1969), "Sampling and testing


of soft rock with weak layers", Geotechnique, V19, p133

5 Evdokimov, P.D., and Sapegin, D.D. (1967), "Stability, shear, and


sliding resistance and deformation of rock foundations", (Israel Prog-
ram for Scientific Translations, Jerusalem)

5 Evdokimov, P.D., and Sapegin, D.D. (1970), "Large scale field


shear test on rock", Proc. 2nd Congo ISRM, Belgrade, V2, paper 3-17

5,8 Fairhurst, C. (1964), "On the validity of the Brazilian test


for brittle materials, IJRM&MS 1 V1, p535

4 Fanshawe, H.G., and Watkins, M.D. (1971), "A sparker survey used
to select a marine dam site in Hong Kong", Qtly J. of Eng. Geol, V4,p25
References 427

2 Farran, J. (1950), "Etude petrographique du comport e ment des


roches dans leurs applications en Genie Civil" Bull. Soci ete Ris t . Na t.
Toulouse, V85, p331

5 Fecker, E. (1970), "Geologische Kartierung des Geb ietes nordwest-


lich von Neustadt /Weinstrosse sowie Bau und Anwendung eines Profilo-
graphen", Diplomarbeit Univ. Karlsruhe (Inst. fur Bo de nme chani k und
Felsmechanik)

2,5 Fecker, E., and Rengers, N. (1971), "Measurement of l a rge scale


roughness of rock planes by means of profilograph and geol o gical
compass", Proc. Int. Symp. on Rock Fracture, Nancy (ISRM) paper 1r1 8

2 Feld, J. (1966), "Rock as an engineering mate ri al" (So i l Test)

3 Fisher, R.A. (1953), "Dispersion on a sphere" Proc . Roy a l So c.


London, Sere A, V217, p295

2 Fookes, P.G., Dearman, W.R., and Franklin, J. A. ( 1971 ) , " Some


engineering aspects of rock weathering with field examples from Dart-
moor and elsewhere" Qtly J. of Eng . Geol . , V4, n3

2 Fookes, P.G., and Rorswill, P. (1970), "Discus s i on OI engi nee ri ng


grade zones", Pro c. Conf. on In-situ TeStin g of Sbils and Ro ck, ( In st.
of Civil Eng., Lon d on) p53

2 Fookes, P. G. , and Parrish, D. G. (1969), "Observation s o n small


scale structural discontinuities in the London Clay and the ir relati on -
ship to regional geology" Qtly J. of Eng . Geol., VI, p217
2 Franklin, J.A. (1970), "Observations and tests for eng i neer ing
description and mapping of rocks", Pro c . 2nd Cong ISRM, Belgrade, VI ,
paper 1-3

2 Franklin, J. A., Brach, E., and Walton, G . (1971 ) "Lo gg ing the
J

mechanical character of rock", Trans. I MM , Sect. A, V81 , p43

2 Franklin, J.A., and Chandra, R. (1972), "The slake durab i lity


test", IJRM&MS, V9, p325

7 Fumagalli, E. (1968), "Model simulation of rock mech a ni cs pr ob -


lems" in Rock Mechanics in Eng. Practise, Zienkiewicz and Stagg , ed.,
(John Wiley)

7 Fumagalli, E. (1973), "Statical and geomechani c al mode ls"


(Springer Verlag)

8 Gale, J. (1975), "A numerical, field, and labora to ry st udy o f


flow in rocks with deformable fractures", PhD Thesis, Un iv. o f Califor-
nia, Berkeley

6 Gallico, A., and John, K.W. (1974), "Graphi c a l s tabili t y an aly ses
for rock abutments of an arch dam" Proc. 3rd Congo ISRM , Denver, V2 B ,
p884
428 References

2 Gamble, J. (1971), Durability - plasticity classification of


shales and other argillaceous rocks" PhD Thesis, Univ. of Illinois

4 Gardner, M. E., and Johnson, C. G. (1971), "Engineering geologic


maps for regional planning"in Environmental Planning and Geology
(U.S.G.S. Off. of Res. ,and Tech., Washington, D.C.) p154

2,4 Geological Society of London Engineering Group Working Party


(1970), "The logging of rock cores for engineering purposes", Qtly J.
Eng. Geol. V3, p1

4 Geological Society of London Engineering Group Working Party


(1972), "The preparation of maps and plans in terms of engineering
geology" Qtly J. Eng. Geol, V5, n4 (W.R. Dearman, P.G. Fookes,
E.G. Smith, and others)
4 Geyer, R. L., and Myung, J. I. (1971), "The 3-D velocity log: a
tool for in-situ determination of the elastic moduli of rocks", Proc.
12th Symp. on Rock Mech. (AIME)

8 Ghaboussi, J., Wi Is on , E., and Isenberg, J. (1973), "Fini te


element analysis for rock joints and interfaces", J.SM&FD ASCE, V99,
p833

8 Golden, A.L., and Troitsky, A.P. (1974), "Behaviour of the rock


mass at the site of a rock fill dam", Proc. 3rd Congo ISRM, Denver,
V2B, p896

6 Goodman, L.E., and Warner, W.R. (1963), "Statics" (Wadsworth,


Belmont, Cal.)

8 Goodman, R. E. (1966), "On the distribution of st resses around


tunnels in non-homogeneous rocks", Proc. 1st Congo ISRM, Lisbon
V2, p249

5,8 Goodman, R.E. (1970), "The deformability of joints", in Determin-


ation of the In-situ Modulus o£ Deformation of Rock - A.S.T.M. STP 477,
p174

5,7 Goodman, R.E. (1972), "Geological investigations to evaluate


stability", Proc. 2nd Int. Conf. on Stability for Open Pit Mines,
Vancouver (AIME)

5 Goodman, R.E. (1974), "The mechanical properties of joints",


Proc. 3rd Congo ISRM, Denver, V1A, p127

5,8 Goodman, R.E. and Dubois, J. (1972), "Duplication of dilatancy


in analysis of join ted rocks", J. SM&FD, ASCE V98, p 399

1 Goodman, R.E. and Duncan, J.M. (1971), "The role of structure


and solid mechanics in the design of surface and underground excava-
tions in rock", Proc. Conf. on Structure, Solid Mechanics, and Engineer-
ing Design, Part 2, paper 105, p1379 (John Wiley)
References 429

5 Goodman, R.E. and Ohnishi, Y. (1973), "Undrained shear testi ng of


jointed rock", Rock Mechanics, V5, p 129

8 Goodman, R.E. and St. John, C. "Stat i c finite element analysis


of jointed rock", in Numerical Methods 'i n Geotech. Eng. ,Christian an d
Desai, eels. rMcGraw-Hill), iIi press

6 Goodman, R.E., and Seed, H.B. (1966), "Earthquake in duced


displacements in sand embankments'!, J .SM&FD ASCE, V92 , p12 5

6,8 Goodman, R .E . and Taylor, R.L. (1967), "Methods of analysis for


ro ck slopes and ab u tments: a review of recent deve lopments" Proc. 8th
Sy mp.on Rock Mechanics (AIME), p303

8 Goodman, R.E. Taylor, R.L., and Brek ke, T. L . (1968), "A model for
the mechanics of jointed r ock ", J .SMFD ASCE, V94 , p637

4 Goodman, R.E., Van, T.K., and Heuze, F. (1972), The measurement


of rock deformabili ty in boreholes", Proc. 10th Symp. on Ro ck Mech. ,
(AIME), p523

4 Gr ant, F. S., and West, G. F. (1 96 5), "Int e rpretation theory in


applied geophysics", (McGraw-Hill)

2 Griffith, J.H. (1937), "Physical properties of typical American


rocks", Bull. Iowa Eng. Experiment Sta. n131, p43

4 Griffiths, D. H., and King, R. F. (1965), "Applied geophysics for


engineers and geologists" (Pergamon Press)

5 Grishin, M.M., and Evdokimov, P.D. (1961), "Shear strength of


structures built on rock", Proc. 5th Int. Conf. on Soil Mech. and
Found. Eng., Paris, p649

8 Grob, H., Kovari, K., and Vannotti. (1970), "Practical applica-


tions of the finite element method to the stability of tunnels" in
German, Proc. 2nd Congo ISRM, Belgrade, V2, paper 4-69

4 Guyod, H., and Shane, L.E. (1969), "Geophysical well logging,


Volume I" (Hubert Guyod, Houston, Texas)

2 Habib, P., and Bernaix, J. (1966), "La fissuration des roches",


Proc. 1st Congo ISRM, Lisbon, V1, p185

4 Hackman, R.J. (1956), "The stereoslope comparator as an instru-


ment for measuring angles of slopes in stereoscopic models", Photo-
grammetric Eng., V22, n5

2 Hamrol, A. (1961), "A quantit at ive classification of the weather-


ing and weatherability of rocks", Proc. 5th Int. Conf. on Soil Mech.
and Foundation Eng, Paris, V2, p771

8 Han, Chin Yuen (1972), "The technique of obtaining equipotential


430 References

lines of groundwater flow in slopes using electrically conductive


paper", MSc Thesis, Imperial College, London.

5 Handin, J., and Stearns, D. W. (1964), flSliding friction of rock",


Trans. Amer. Geophysical Union, V45, p103

4 Harbaugh, J.W., Bonham-Carter, G., Graeme, (1970), "Computer


simulation in geology" (Wiley-Interscience)

5 Haverland, M. L., and Slebir, E. J ., (1972), "Methods of performing


and interpreting in-situ shear tests fl , Proc. 13th Symp. on Rock Mech.,
(ASCE), p107

2 Hawkes, I., and Me llor, M. (1970), "Uniaxial testing in rock


mechanics laboratories", Engineering Geology, V.4, n3

5 Hayashi, M. (1966), "Strength and dilatancy of brittle, jointed


(rock) masses - - the extreme value stochastics and anisotropic
failure mechanism", Proc. 1st Congo ISRM, Lisbon, V1, p295

5 Hayashi, M., and Fujiwara, Y. (1968), "A mechanism of anisotropic


dilatancy and shear failure of laminately jointed rock masses fl , Tech.
Rep. C67006, Cent ral Res. Inst. of Electric Power Industry, Toky-o--

8 Hayashi, M., and Hibino, S.(1968), "Progressive relaxation of


rock masses during excavation of an underground cavity", Proc. Int.
Symp. on Rock Mechanics, Madrid (ISRM)

5 Hayashi, M., and Kitihara, Y. (1970), "Anisotropic dilatancy and


strength of jointed rock masses, and stress distribution in fissured
rock masses", in Ro'ck Mechanics in Japan, VI, p85 (Japan Soc. of Civil
Engineers)

4 Heiland, C.A. (1946), flGeophysical exploration" (Prentice Hall)

6 Hendron, A.J. Jr., Cording, E.J., and Aiyer, A.K. (1971),


"Analytical and graphical methods for the analysis of slopes in rock
masses", U.S.Army Corps of Engineers, Nuclear Cratering Group Rep 36

8 Heuze, F.E., and Goodman, F.E. (1967), "Mechanical properties


and in-situ behavior of the Chino limestone, Crestmore mine, Riverside
Cal.", Proc. 9th Symp. on Rock Mechanics,(AIME)

8 Heuze, F.E., and Goodman, R.E. (1970), "The design of room and
pillar structures in competent, jointed rock, example: the Crestmore
mine, Cal.", Proc. 2nd Congo ISRM, V2, paper 4-41

8 Heuze, F. E., Goodman, R. E., and Bornstein, A. (1971), "Numerical


analyses of deformability tests in jointed rock", Rock Mech. V3, p13

6 Heuze, F.E., and Goodman, R.E. (1972),"Three dimensional approach


for design of cuts in jointed rock", Proc. 13th Symp. on Rock Mechanics
(ASCE), p347
Re ferences 431

3,4 Higgs, D. V ., a nd Tunell , G. (1959), " Angul a r r e lati o ns of l in es


and p l anes " , (Wm C. Brown , Dubuque , Iowa )

2,5 Hodgs on , R.A. (196 1 ), IIClassification of s t ru ctures o n joi n t


sur f aces II, Amer. J. of Sci. V 259 , p493

6,8 Hoeg, K. , a n d Muraka , R. (1974 ), fl Probab i list i c a n a lysi s an d


design o f a retai n ing wall ", J. Ge o te ch. Eng. Div., Pro c . ASeE, V100,
p3 49

6 Hoek , E . ( 1 970), "Estimating t he stability o f e xcava t ed sl opes


in open cast mines Trans. I.M.M., Sect. A, V79 , pAI09
lf
,

1 Hoe k, E . (1971), I'Rock e ngineering" , Innaugura l lect ure, I mp e ri a l


Co lle g e , Lo ndon

1 Hoek , E., and Bray , J. ( 1974 ), " Rock s l ope e n g i nee ring II (I MM,Lon )

2 Ho ek , E. and Fran k l in , J.A. (1968) , "A simp le tr iaxial c el l fo r


f ie ld or laborator y test ing of r ock" , Trans . IMM , Se ct .A , p2 2

6 Hoek , E., Br ay , J., and Boyd , J. (1973), liTh e stabili t y o f a ro ck


s lope con t a i ning a we dge res ti ng on t wo intersecting d iscont in ui t ie s" ,
Qtl y J. Eng. Geol ., V6 , p1

5 Hoek , E. , and Pentz, D .L. (1968), " Tbe stab i l i ty of op e n pit


mi n e s ", Rock Mech. Re s . Rep. n5, Imperial Colle ge , London

8 Ho f man , H. (1970), " The d e formation proce s s of a r e gula rl y join t ed


discontinuum during the excavation o f a cut" (i n Ge r man), Pro c . 2nd
Congo ISRM , Be lgr ade , V3 , paper 7-1

4 Ho ll ingsworth , J. (1974), "Two dimension al st ructura l de fe ct


analysis t e c h n i ques ", CSIRO (Australia, Div. of Appl. Geome ch . Tech .
Rep . 21

5 Hor i no I F . G. , Hoskins , J. R. 1 and El lickson J M. L. (1968 ) , "A


me t hod o f meas uring surface te x t ur e o f rock ", U .S. B . Mine s R.I. 7095

5 Horn, H. M. , and Deere , D. U . (1962), "F r ict iona l c haract eristi cs


of mine r als ", Geot echnique, V12, p319

5 Hoskins, E . R., Jaeger, J. C ., and Ros engre n, K. J. (19 68 ) , " A


me d i um scal e direct friction e xperimen t" IJRM&MS , V5 , p143

8 Hoyaux, B. , and Ladanyi , B . ( 1970) , "Gravitati on a l sre ss fiel d


aroun d a t unne l in s of t ground" , Can. Geot e ch. Jour ., V7, p54

2 rida, R ., and oth ers (1970), "Geologica l rock c l assi fi c ation of


darn fo u ndations" , i n Rock Me ch. in Japan, V.1 , p171 ( J apan Soc . Civ En g)

5,8 Iida , R . , a n d Kobay ash i , S . ( 1974) ," The o ret i ca l s t u d y o n t h e non-


e lastic b ehav i or of join te d rock mas se s " Jour . of Res e arc h, Public
432 References

Works Res. Inst. I Ministry of Construction I V16 I p35 (in English)

5 Iida , R., Hojo, K, and Harada, J. (1974), "In-situ tests and


theoretical studies on the relations between looseness and deformation
characteristics of jointed rock masses" Proc. 3rd Congo of ISRM, Denver,
V2B, p719

2 Iliev, I.G. (1966), "An attempt to estimate the degree of weath-


ering of intrusi ve rocks from their physico-mechanical properties",
Proc. 1st Congo ISRM, Lisbon, V1, p109

2 Ingram, R.L. (1954), "Terminology for the thickness of stratifi-


cation and parting units in sedimentary rocks", Bull. G.S.A., V65, p937

3 Irving, E. (1964), "Paleomagnetism and its applications to


geological and geophysical problems" (John Wiley)

8 Isenberg, J. (1972), "Analytic modelling of rock structure


interaction", Report from Agbabian Assoc. to U.S.B.Mines, (AD 749-373)

4,7 Jack, H. (1969), "Tunnel mapping methods, a discussion of a


paper by Cooper" BUll. A. E. G., V6, p151

2 Jacquet, P. A., and Mencarelli, E. (1959), "Technique non-destruc-


tive pour l'etude des surfaces de cassure au microscope electronique",
Compte Rendu Acad. Sci. Paris, V248, p2477

1 Jaeger, C. (1972), "Rock mechanics and engineering" (Cambridge


Univ. Press)

5 Jaeger, J.C. (1959), "The frictional properties of joints in


rock", Geofisica Pur a e Applicata, V 43, p148

5 Jaeger, J. C. (1960), "Shear fracture of anisotropic rocks",


Geol. Magazine, V97, p65

5 Jaeger, J .C. (1971), "Friction of rocks and the stability of


rock slopes - Rankine Lecture", Geotechnique, V21, p97

1 Jaeger, J.C., and Cook, N.G.W. (1969), "Fundamentals of rock


mechanics", (Methuen)

5 Jaeger, J.C. and Rosengren, K.J. (1969), "Friction and sliding


of joints" Proc. Australasian Inst. of Min. and Metallurgy, n229, p93

4 Jakosky, J.J. (1950), "Exploration geophysics" 2nd edition,


(Trij a)
4 Janke, N.C. (1972), "Field measurements with common equipment",
Photogrammetric Eng., V38, p37

3 Jeran, P. W., and Mashey, J. R. (1970), "A computer program for


the stereographic analysis of coal fractures and cleats", U.S.B.Mines
Inf. Cire. 8454
References 433

2,6 John, K. W. (19 62 ), "An appro ach to rock mechan i cs " J SM&FD ASCE
V88, p1

6 John, K. W. (196 8) "Gr ap h ic a l s t ab i l ity a nal ysis of slop e s in


I

jointed rock", J.SMFD , ASCE, V9 4, n SM2

5,7 John, K.W. (1970), "Civil engineeri ng a p p r oach t o e v aluate


strength of a regularly jointed rock ", Pro c 11th Syrnp. on Rock Mech.
(AIME), p69

6 John, K.W . ( 1 9 70) , "Three dime nsion al s t a b ility ana l ysi s o f sl o pes
in jointed rock" Pr o c . Johan ne sb u r g Syrnp o n Ro ck Mech , South Africa

5 Jouanna, P. (19 72 ), "Essais de p e r c ol at i on a u laboratoire sur


des echantillons de micaschiste soumis a des c ont r ai nt e s f l , Proc.
Symp. on Per co latio n t hrough Fiss ure d Rock, Stut tgar t, ( ISRM) p aper T2-F

5,7 Kandau rov , I. I., Ouva r ov, L.A. , a nd Karpov , N .M . ( 1974) , I f Con-
trainte dans les mo deles des mas s if s r o che ux fi s sure s s ans pousse e
horizontale" Pro c . 3rd Cong o ISRM,De nve r, V2A, p157

8 Kawamoto , T . ( 1970) , "On t h e st a t es o f st r e s s and deformat ion


of natural s l opes", in Rock Me chanics i n J a pan, (Japan Soc . Ci v. En g) ,
Vl, p85

4 Keller, G. V. , a nd F r i schkn echt, F. C . (1970) , "Elec t rica l methods


in geophysical p r ospec t ing " ( Perg amon )

2 Kiersch, G .A., a nd Tre as he r, R . C . (1955) " In ves t igat ions , a eri a l


and engineering ge ol ogy - Folsom Dam proj e ct , Cen tr al Ca l " Econ. Ge ol .
V50, p271

4 Kiraly, L. (19 69 ), "Statis tical ana lysis of f rac t ures (orientation


and densi ty)" Sonder dru ck aus de r Geologi sche n Run dsch a u, Band 5 9 , p12 5

3 Kirschke , D. ( 1970 ), "Die p ro g r ammie r s p rac he GELI 1 in lbrt r an IV


Version fur di e Re c henlage IBM 36 0" Cl aus t hale r Te kt o n isch e Hef t e 10,
p133

4 Kneuper, G. et al. (196 7 ), " New re sults wit h a refl e ct i on se ismi c


method to locate tectonic faults in coa l mini n g" Bergbauwissensch a ft en
V14, p428

2 Komarni t skii, N.N . (1 9 68) , "Zon e s an d plan e s of we akne s s in r ocks


and slope stability" ( Consultants Bure au, N.Y . )

7 Korbin, G.E. (1975), "A model s tudy of s pi l ing rei n f o rcemen t in


underground openings" PhD Thesis, Univ. of Ca l. Berkel ey

4 Krumbein, w.e., and Graybi ll, F.A. ( 1 9 6 5), !IAn i ntroduct ion to
statistical models in geolo gy" (M cGraw-Hill )

5 Krsmanovic, D. (1967 ) , " In iti al a nd res i dua l shear s treng th o f


hard rocks", Geotechnique, V1 7 , p14 5
434 References

7 Krsmanovic, D. (1971), liOn the results of measurement of stresses


and strains in the rock mass of a geostatical model of the Grancarevo
Dam ll Publ . . of the Inst. of Geotechnicsand Found. Eng., Faculty of
Civil Engineering, Sarajevo, Yugoslavia
5 Krsmanovic, D. and Langof, Z. (1964), IILarge scale laboratory
tests of the shear strength of rocky material ll Rock Mech. and Eng. Geol.
Supplement II, p20
7 Krsmanovic, D., and Langof, Z. (1971), "On the results of testing
the Geostatical model of the Grancarevo Dam"(see Krsmanovic, D. (1971»
5 Krsmanovic, D., and Popovic, M. (1966), "Large scale tests of
the shear strength of limestone", Proc. 1st Congo ISRM, Lisbon, V1,p773
8 Kruse, G.H. (1971), "Power plant chamber under Oroville Damll,
Proc. Symposium on Underground Rock Chambers (ASCE) p333
2 Kruse, G.H., Zerneke, K.L., Scott, J.B., Johnson, W.S., and
Nelson, J. S. (1970), IIApproach to classifying rock for tunnel liner
design II, Proc. 11th Symp. on Rock Mech. (AIME), p169
1 Krynine, D., and Judd, W. (1959), IIPrinciples of engineering
geology and geotechnics" (McGraw-Hill)
8 Kulhawy, F.H., and Duncan, J.M. (1972), "Stresses and movements
in Oroville Damll, JSM&FD ASCE, V98, p653
5 Kutter, H.K. (1971), IIStress distribution in direct shear test
samples", Proc. Int. Symp. on Rock Fracture, Nancy, (ISRM), paper 2-6
5 Ladanyi, B., and Archambault, G. (1970), IISimulation of shear
behaviour of a jointed rock mass ll Proc. 11th Symp. on Rock Mechanics,
(AIME) p105
5 Ladanyi, B., and Archambault G. (1972), "Evaluation de la resis-
tance au cisaillement d'un massif rocheux fragmente" Proc. 24th Int.
Geol. Congo Montreal, Section 13, p249
7 Lajtai, E.Z. (1967), "The influence of interlocking rock discon-
tinuities on compressive strength (model experiments)" Rock Mech. and
Eng. Geol V5, n4 p217
5 Lajtai, E.Z. (1969a), "Shear strength of weakness planes in rock"
IJRM&MS, V6 p499
5 Lajtai, E. Z. (1969b), "Mechanics of second order faults and
tension gashes", Bull. G. S. A., V80, p2253
5 Lane, K.S., and Heck, W.J. (1964), "Triaxial testing for strength
of rock joints" Proc. 6th Symp. on Rock Mech., Univ. of Missouri, p98
8 Lane, R.G. (1970), "An investigation into the deformation of
a combined dam and powerhouse structure" Proc. 2nd Congo ISRM, Belgrade
References 436

V3, paper 8- 2

7 Langh aar , H.L. (1951), "Dimensional analysis and theory of models"


(John Wiley)

4 Lattman, L. R . , a nd Ray, R.G. (1965), "Aerial photographs in


field geology" (Holt, Rinehart and Winston)

2 Lauffer, H. (1958), "Gebirgsklassifizierung fur den Sto l lenbau"


Geologie und Bauwesen, V24, p46

4 Lauterbach, R. (1953), "Mikromagnetik, ein Hilfsmittel ge olo-


gischer Erkundung" Wis s . Z. Karl Marx Vniv. Leipzig, V3, p223

8 Lee, K.L., a n d Shen, C.K. (1969), "Horizontal movements rel ated


to subsidence" J SMFD ASCE V95, p139

4,2 Leet, L.D., and Judson, S.(1971), "Physical geology, 4th edi t i on"
(Prentice-Hall)

4 Library of Cong ress Geography and Map Division (1973), "The


bibl i ography of c art o gra phy" (G.K.Hall and Co.)

4 Linkwi tz, K. (1963), "Terrestrisch photogrammetris che Kl uft mes-


sung" Rock Mech. and Eng. Geol. V1, p153

5 Locher , H.G . ( 1 968), "Some results of direct shear t e sts o n


rock discont inuit ies ", Proc. 2nd Int. Symp on Rock Me ch., Madri d ( ISRM)
V171

3,5 Lokin, P. ( 1974), "Discontinuity anisotropy of rock masses ", Pro c.


3rd Congo I SRM , Denver V2A, p174

5 Lombar di , G. and Dal Vesco, E. (1966), "Die experimentel le


Bestimmung der Reibung skoeffizien t e n fur die Felswiderlager de r Sta u-
ma ue r Con t r a (Verzasca)" Proc. 1st Cong o ISRM Lisbon, V1, p571

6 Londe , P. ( 1965 ), "Vne me thode d' an a lyse a


trois dime n siona de la
stabili te d' une ri ve r o che use " Annales' des Ponts et Chaussee s, nl, p37

1 Londe , P. (196 7), "La mecanique des roches - - discipline nouv-


elle au service de la construction des gran ds barrages" Sci en ce Prog r es
- La Nature , n3384, April (Dunod, Paris)

6 Londe, P., Vi gi er, G., and Vormeringer, R. (1969), "St ability of


rock slopes, a three d imensional study" J SM&FD ASCE, V95,nSM7

6 Londe, P., Vigier, G., and Vormeringer, R. (1970), IIStabili t y of


rock slopes, graphical methods" J SM&FD ASCE V96, nSM4

2 Loughn an , F . C. (1969), IIChemi c a l we a thering of t he silica te


minerals" (Elsevier)

4 Louis, C., and Maini, T. (1970), "The determination o f hydrauli c


Goodman-Geological En<j .-29
436 References

parameters in jointed rock" Proc. ~ 2nd Gong·. TSRM, Belgrade, V1,


paper 1-32

4 Lugeon, M. (1933), "Barrages et geologie", (Dunod, Paris)

2 Lumb, P. (1962), "The properties of decomposed granite",


Geotechnique, V12, p226

3 Mahtab, M.A., Bolstad, D.D., Alldredge, J.R., and Shanley, R.J.


(1972), "Analysis of fracture orientations for input to structural
models of discontinuous rock", U. S.B.MInes R.T.7669

8 Mahtab, M.A., and Goodman, R.E. (1970), "Three dimensional finite


element analysis of jointed rock slopes", Proc·.2nd Congo ISRM, Bel-
grade, V3, paper 7-12

4 Maini, Y.N.T. (1971), "In-situ hydraulic parameters in jointed


rock, their measure"ment and interpretation", PhD Thesis, Imperial
College, London

8 Malina, H. (1969), "The numerical derermination of stresses and


deformations -in rock tak~ng into -account discontinuities" Proc. 19th
ColI. on Geomech., Salzburg

8 Markland, J .T., (1972), "A useful technique for estimating the


stability of rock slopes when the rigid wedge slide type of failure
is expected" Imperial College Rock Mech. Res. Rep. n19

2,5 Martin, G.R., and Millar, P.J. (1974), "Joint strength character-
istics of a weathered rock", Proc. 3rd Congo ISRM, Denver, V2A, p263

5 Mathews, K.E. (1970), "Excavation design in hard and fractured


rock at the Mount Isa Mine, Australia", MSc. Thesis, Univ. of Queensl.

4 McGregor, K. (1967), "The drilling of rock", (C.R.Books Ltd,


London)
2 McKee, E.D., and Weir, G.W. (1953), "Terminology for stratifica-
tion and cross stratification in sedimentary rocks" Bull. G.B.A., V64,
p381

5 McLamore, R., and Gray, K.E. (1967), "The mechanical behavior of


anisotropic sedimentary rocks", Trans. A.S.M.E., p62

3 McMahon, B. K;. (1967), "Indices related to the mechanical proper-


ties of jointed rock", Proc. 9th Symp. on Rock Mech. (AIME), P117

4 Meissner, R. (1961), "Wave front diagrams from uphole shooting",


Geophysical Prospecting, V9, p40
References 437

4 Meke l, J.F. , Savage , J. F. , a nd Zo rn, R.C. (1964), "Slope meas u re -


ments an d es t i ma t i ons from aerial photos", Int. Training Cent re fo r
Aer i al Survey (ITe ) , Delft , pub l. Seri es B, ri26

5 Men el, V. (1965 ), "Di la t ancy of rocks", Rock Me ch . and Eng. Geol.
V3, p5 8

4 Meri fi eld, P .M ., et. al. (1969 ), "Satellite imag ery of the e arth"
Ph otogrammetr ic Eng . V35, p654

4 Moffit, F .R. ( 1959), "Photograrrunetry", (International Textbo ok)

4 Moll ard, J.D. (196 2 ), "Photo a na l ys is and interpretation in


engineering ge ology in v estiga tions, a re view" Revi e ws in Engineer i ng
Geology (G.S.A. ) , F l uhr an d Leggett, eds. V1

8 Morgenstern, N.R., an d Guthe r, H. ( 1972), "Seepage into an excav-


ation in a medium possessing st re ss d ependent permeability", Symp . o n
Percolation Throug h Fissured Rock, Stuttgart (ISRM ), paper T2-C

5 Morlan d , L. W. ( 1 974 ), " Cont inuum model of regularly jointed


me diums", J. Geop h. Res . , V79, p35 7

2 Mor l ier , P. ( 1968 ), "Rel ation e ntre la fissuration et la ce lerit e


des ondes dans les r oches fissurees", Co Tloq ue sur l a F i ssur a t i on des
Roches - Revue de L ' I ndus trie Mi ne'r a l, ' h lim . sp. 15 May , p .16

2,4 Moye, D. G. (1 9 5 5 ), "Engineerin g ge ology for the Snowy Mtn. scheme


Australia" J . lnst . of En g inee r s, Au s trali a, V27, p287

4 Moye, D.G. ( 1 96 7), "Diamon d dri ll ing for foundati on exp l or a t ion" ,
Trans. Inst . o f Enginee rs, Au stra li a, V9, p 95

1 Miiller, L. (1 96 3), "Der fel sbau", volume 1 (F.Enke, Stuttgart)

1 MUl le r, L. ( 1 96 4), "The rock slide in Vajont Va lle y", Rock


Me ch. and Eng . Geo l. , V2, p 148

7 Miiller, L. (1 9 66), "Der progressive Bruch in gekluf t eten Median "


Proc. 1 s t Congo ISRM, Lisbon, V1, p679

1 Miiller, L. (1968), "New consi de rations on the Vajont slide",


Rock Mech. and En g . Geol., V6, p1

2 Mi.iller ; L., a n d Hofmann, R. (1970), " Se lection, compil a tion, and


as sessment of geological d ata for the slope problem" P roc. Symp. on
Planning Open Pit Min e s, Joh annesbur g , S. Afr.

7 Miiller J L., and Pacher, F. (1965), "Modell versuc he fur Klarun g


der Bruchge fahr g ek lufteter Medien", Rock Me ch. and Eng. Geol. Sup p.I I,
p7

4 Musgrave, A . W. ( 1 967), "Seismic refraction prospecting';, (Soc.


438 References

Expl. Geoph., Tulsa)

4 Myung, J., and Bal tosser, R. W. (1972), "Fracture evaluation by


the borehole logging method", Proc. 13th Symp. on Rock Mech. (ASCE)

2 Nascimento, U. (1970), "0 problema da al terabilidades das rochas


em engenharia civil", Revista Fomento, V8, n2

5 Nascimento, U., and Teixeira, H. (1971), "Mechanisms of internal


friction in soils and rocks", Proc. Symp. on Rock Fracture, Nancy
(ISRM), paper 2-3

6 Newmark, N. (1965), "Effects of earthquakes on dams and embank-


ments", Geotechnique, V15, n140

2 Nickelson, R.P., and Hough, V.N.D. (1967), "Jointing in the


Appalachian plateau of Pennsylvania" BUll. G.S.A., V78, p609

8 Noorishad, J., Witherspoon, P. A., and Maini, Y. N. T. (1972), "The


influence of fluid injection on the state of stress in the earth's
crust", Proc. Symp. on Percolation Through Fissured Rocks, Stuttgart,
(ISRM), paper T2-H

2 Nur, A. and Simons, G. (1970), "The origin of small cracks in


igneous rocks", IJRM&MS, V7, p307

1 Obert, L., and Duvall, W. (1967), "Rock ~echanics and the design
of structures in rock" (John Wiley)

2 OIlier, C. (1969), "Weathering" (Elsevier)

5 Ohnishi, Y. (1973), "Lab. measurement of induced water pressures


in jointed rocks", PhD Thesis, Univ. of California, Berkeley

2 Onodera, T. F. (1963), "Dynamic investigation of foundation rocks


in-s i tu", Proc. 5th Symp. on Rock Mech. (Macmillan), p517

2 Onodera, T.F. (1970), "Activities in rock mechanics in the Japa-


nese Society of Soil Mechanics and Foundation Eng." in Rock Mech. in
Japan, V1 (Jap. Soc. Civ. Eng.)

8 Oudenhoven, T.F., Babcock, C.O., and Blake, W. (1972), "A method


, for the prediction of stresses in an isotropic inclusion or orebody
of irregular shape", U.S.B.Mines, R.I. 7645

8 Pariseau, W.G., and Stout, K. (1972), "Open pit mine stability:


the Berkeley pit" Proc. 13th Symp. on Rock Mech. (ASCE), p367

8 Pariseau, W.G., Voight, B., and Dahl, H.D. (1970), "Finite


element analyses of elastic plastic problems in the mechanics of geo-
logic media", Proc. 2nd Congo ISRM, Belgrade, V2, paper 3-45

5 Paterson, M.S., and Weiss, L.E. (1966), "Experimental deformation


and folding in phyllite", Bull. G.S.A., V77, p343
References 439

5 Pa tton , F.D. (1966), "Mul tiple modes of sh e ar fai lure in rock" ,


Proc. 1st Con g o I SRM, Lisbon, V1, p5 09

5 Pau ldin g , B. W. (1 970 ), "Coef fic ient of fr i c t ion of nat ural ro ck


surf ace s" , J .SM&FD ASCE , V96m nSM2

2 Pender , M. J. ( 1 971), "Some p rope r ties o f weathe re d graywa c ke"


P r oc . 1st Aust ra l ia-New Zealand Con f. on Geome c hanics, V1, p423

3 Phi l lip s, F. C. (1 946 ), "An intro duction to cryst al lography",


(Longmans, Green)

3 Phi l lips , F . C . (1 97 1 ) , "Th e use of st ereogr aphi c proj e ct ion in


structural geology, 3rd edi t ion" ( Edward Arnol d )

8 Phuk a n, T.A ., Lo, K., and La Roche l le , P. ( 1 969) , " Stress e s an d


deformations of vertical slopes in e las to - p l a stic rocks ", P roc. 1 1 th
Symp. on Rock Mech . (AIME), p19 3

3 Pincus , H. J. (1953 ), "The analys is of agg regates o f orient a tion


dat a in the earth sciences", J. o f Ge ol., V61 , p 4 82

4 Pincus, H.J. (19 73 ), "A modified pocket t ransi t f o r me asuring


strike on the un de rs i de of surfaces" , I J RM&M S, V10 , p B 3

2,5 Piteau, D.R. (1973), "Charact eriz in g and ext rapo lating rock
joint properti es in e ng ineering practi c e", Ro ck Mech. Sup p . II, p5

4 Polak, E.J. (1963), "The measurements o f, re l at ion betwe en, and


factors affecting the properties of rocks", Proc. 4 th Aust ra li a , Ne w
Zealan d Conf . o n Soil Mech. an d Found . En g . , p220

4 Pol ak , E.J. ( 19 71), "Seismic a ttenuation in e ngi nee r ing si t e


investigations", P r o c. 1st Austra lia, New Zeal"an d Conf. on Geome ch a n i c s
Melbourne, V1, p430

5 Prat t , H. R., Black , A.D., and Brace, W.F. ( 1974), "Friction an d


deformation of j o in te d quartz diorit e ", Pro c. 3rd Cong o I SRM, Denve r,
V2A, p306

2,4 Pri ce, N. (1 966 ), "Fault a nd j o int de velopment in b ri tt le and


semi-brit t l e ro ck " ( Pergamon Press)

4 Ragan , D .M. (197 3), " St r uc tura l geology - - a n i n troduction to


geometrica l techniques , 2nd edition", ( J ohn Wi le y)

5 Rale i gh , C.B., and Pate rson, M.N. (1 96 5), " Exp erime ntal deforma-
tion of serpe ntine and i ts tectonic i mplica t ion s" , J . Geoph . Res .
V70, p396 5

4 Ramsay, J.G. ( 1967), "Fo l ding an d fract uri ng o f ro cks" ( McG r a w-


Hill)

4 Ray, H.G. ( 1960), "Aer ial pho tog r aphs in geo lo g ic i n t erpret at ion
440 References

and mapping", U. S. G. S. Prof. Paper 373

2 ReichmutJ?., D.R. (1963), "Correlation of force-displacement data


with physical properties of rock for percussive drilling systems",
Proc. 5th Symp. on Rock Mech. (Pergamon), p49

4 Rengers, N. (1967), "Terrestrial photogrammetry: a valuable tool


for engineering geological purposes" Rock Mech. and Eng. Geol., V5,
p150

2,5 Rengers, N. (1970), "The influence of surface roughness on the


friction properties of rock planes", Proc. 2nd Congo ISRM, Belgrade,
V1, paper 1-31

5 Rengers, N. (1971), "Unebenhei t und Reibungswiderstand von


Gesteinstrennflachen", Veroffentli"chunge"n des Inst. fur Bodenmechanik
und Felsmechanik der Uni v. Frederic"anna" in Ka"rlsruhe, V47

8 Rescher, O.J. (1968), "Rock reinforcement of the cavern at


Veytaux by rock anghors and pneumatically applied concrete"(in German)
Rock Mech. and Eng. Geol., Supp. 4, p216

8 Reyes, S.F. and Deere, D.U. (1966), "Elastic plastic analysis of


underground openings by the finite elemen t method',' Proc. 1st Cong.
ISRM, Lisbon, V2, p477

2,5 Richards, L.R. (1972), "Classification and weathering of near


surface, jointed rock" MsC Thesis, Imperial College

5 Ri chards, L. P. (1973), "Strength properties of De labole s la tes" ,


Imperial College Rock Mech. Res. Rep.n22

5 Ripley, C.F. and Lee, K.L. (1961), "Sliding friction tests on


sedimentary rock specimens", Trans. 7th Int. Gong". on Large Dams, V4,
p657

4 Robinson, J.B., Brown, N.P., and Benns, F.T. (1967), "Drilling


equipment for difficult coring conditions; new type core lifter and
triple tube core barrel", Proc. Aus"tala"si"an" rnst. o"f Min. and Metal. ,
n224, p37

5 Robinson, L.H., and Holland, W.E. (1969), "Some interpretation


of pore fluid effects in rock failure", Proc. 11th Symp. on Rock
Mech. (AIME), p585

4,5 Rocha, M. (1971), "A new method of integral sampling of rock


masses", Rock Mech. V3, p1

4 Rocha, M. (1974), "Recent possibilities of studying foundations


of concrete dams", Proc. "3rd Congo TSRM, Denver, VlA, p879

4,5 Rocha, M., and Barroso, M. (1971), "Some applications of the


new integral sampling technique", Pr"oc". In"t. Symp. on RO"ck Fracture,
Nancy (ISRM), paper 1-21
References 44 1

4 Roch a, M., DaSilvei ra , A. , et . a l. (1970) , "Characterization of


the deformab ili ty o f rock masses by dilatomete r tests" , Proc . 2nd
Cong o ISRM, Belgrade , V1, p aper 2- 32
8 Roda tz W. , and Wi ttke , W. ( 1972 ), "I nteraction betwe en defor m-
I

a t ion a nd perco l ation in fissured , anisotropic rock" ( in German ), Proc .


Symp. on Perco lat ion Thr ough Fiss ure d Rock, Stuttgart (ISRM) paper T2-I
3 Rodgers J . ( 1952) , "The use of equa l are a or other pro j ect i ons
1

in the st atis t i ca l t reat men t of j Oi nts" , Bu ll. G.S.A., V.63 , p427


5 Romero, S . (1968) , Pl l n - s i t u direct shear t ests on i rregul ar
s ur f ace joints f ill ed wit h clayey materi al " , Proc. Int. Symp. on Rock
Mech., Madri d (I SRM), V1 , p189
7 Rosenb l a d, J . L. (19 70 ), "Failure mo des o f models of joint e d ro ck
masses PI, Pr o c. 2n d Cong. ISRM, Belgrade, V2, paper 3-.11
4 ,5 Ros e n gren, K. J. (196 8 ), "Ro ck mech an i cs of the Black St a r op en
cut , Moun t Isa" PhD Thesis , Aust rali an National Un i v., Canberra
4 _ Rosengren , K . (.1970 ), "Diamond dri ll i ng for struct ura l purposes
at Mt . I sa ',' I ndustria l Diamond Review, p388
4 Ros s - Brown , D. M., and Atk inson , K. B. (1972), "Terrestri al photo-
gr ammetry in open p it s ; p art 1 - des cr iption and use of the pho to
theode li t e in mine surveyin g ", Trans . I MM, Sec-t . A, V81, p205
4 Ross - Brown , D.M. , Wickens , E. H. , and Mark l and , J. ( 1973 ),
"Te rrest r i al photogrammet ry in open pi ts ; part 2 - an aid to geol .
mapping " Trans. IMM, Sect . A, V82 , pll5
4 Roxtrom , F .B. (1961 ), "New core orientati on device " Econ. Geol.
V56, p1 310
4 Roxtrom, F.B. ( 1961), "Cr ael i us aut omat i c core arienter, inclined
dri ll holes ", Can . Min . J ., V82 , p60
5 Rowe , P.W . (1962 ), "The st r ess dil a tancy r elation f or stat ic
equilib ri um of an ass embly of particles i n contact" , Proc . Roy a l Soc.)
London, A. , V269 , p 500
5 Rowe, P. W., Barden , L. , and Lee, I. K. ( 1964), "Ene rgy componen t s
duri ng the triaxi al ce ll and direct shear test ll , Geotechniq ue, V14,
p247
6 Ru iz 1 M.n . ( 1974), "Minimum ancho ri ng cost for stabi li zati on of
ro ck s opes" Proc . 3rd Cong o I SRM, Denver , V2B~ pS13
5 Ruiz, M., an d Camargo, F. (1966) } "A l arge scale f ield tes t on
ro ck " , Pro e. 1st Cong o ISRM , Lisbon , Vl, p25 7
5 Ruiz , M.D ., Camargo , F.P. , Midea , N.F. , and Nieb le, C. M. (1968) .
"Some considerat ions regarding the shear s t re ngth o f rock masses" ,
442 References

Proc. Int. Symp. on Rock Mech., Madrid (ISRM), paper 2-5, p159

2 Ruxton, B.P., and Berry, L. (1957), "Weathering of granite and


associated erosional features in Hong Kong", Bull. G.S.A., V68,- p1263 '

2 Rzhevsky, V., and Novik, G. (1971), "The physics of rocks",


translated by A.K. Chaterjee (Mir publishers, Moscow)

4 Sabarly, F. (1965), "Mesures des permeabili ties en place", a


report prepared by Geoconeil, 31, Route de Versailles, 78170 La CelIe,
Saint Cloud, France

2 Sahores, J. (1962), "Con tribut ion to the study of the mech. phen.
accompanying hydration of anhydrite" (in French), Centre d'etudes et
de Recherche de L'Industries des Liants Hydrauliques, Pub. Tech. n126

8 St. John, C.M. (1971), "Numerical and observational methods of


determining the behaviour of rock slopes in open cast mines", Imperial
College, PhD Thesis.

8 St. John, C. (1972), "Finite element analyses of two and three


dimensional jointed structures - computer programmes" Imperial ColI.
Rock Mech. Res. Rep. n13 and appendix

4 Savage, J.F. (1965), Terrestrial photogrammetry for geol.


purposes", Int. Train. Centre for Aerial Survey (ITe), Delft, publ.
series B, n33, p41

2 Schneider, B. (1967), "Moyens nouveaux de reconnaissance des


massifs rocheux", Annales de l' Ins't. Tech. duo Batiments et des
Travaux Public, Supp, V20, p1055

5 Schneider, H.J. (1974), "Rock friction - a laboratory invest."


Proc. 3rd Congo ISRM, Denver, V2A, p311

2,5 Scott, R.F. (1963), "Prin. of soil mechanics" (Addison-Wesley)

2 Selmer-Olsen, R. and Blinciheim, O.T. (1970), "On the drillability


of rock by percussive drilling", Proc. 2nd Gong. ISRM, Belgrade, V2,
paper 5-8
2 Serafim, J.L. (1964), "Rock mechanics considerations in the
design of concrete dams", in St ate o'f Stress In the Earth's Crust,
Judd, ed. (Elsevier), p611
5 Serafim, J.L. and del Campo, A. (1965), "Interstitial pressures
on rock foundations of dams", J.SM&FD, ASCE, V91, p65

5 Serafim, J.L., and Lopes, J.B. (1961), "In-situ shear tests and
triaxial tests on foundation rocks of concrete dams l l , Proc. 5th Int.
Conf. on Soil. Mech. and Found. Eng., Paris, V1, p533

8 Serata, S. (1970), IIPrerequisites for applic. of finite element


methoq. to solution cavities and conventional mines l l , Proc. 3rd Symp.
References 443

on Salt, Rau and Delwig, eds. (Northern Ohio Geol. Soc., Cleveland)
V2, p249

3 Shanley, R.J., and Mahtab, M.A. (1974), "A computer program for
clustering data points on the sphere", U.S.B.Mines Inf. Circ. 8624

3 Shanley, R.J., and Mahtab, M.A. (1975), "FRACTRAN: a computer


code for analysis of clusters defined on the unit hemisphere" U.S.B.
Mines Inf. Circ. 8671

5 Sharp, J.C., and Maini, Y.N.T. (1972), "Fund. considerations on


the hyd. char. of joints in rock", Proc. Symp. on Perc. Through Fis-
sured Rock, Stuttgart (ISRM)
2 ,4 She r ar d , J. L., C 1 u f f, L . S " an d A11 en, C. R . ( 19 74 ) , "Po t en t i a 11 y
active faults in dam foundations", Geotechnique, V24, p367

4 Simpson, B. (1968), "Geological Maps", (Pergamon)

5 Skempton, A.W. (1954), "The pore pressure coefficients A and B"


Geotechnique, V4, p143

4 Slichter, L.B. (1955), "Geophysics applied to prospecting for


ores" Economic Geology, 50th Anniversary Volume, Part II, p885

4,5 Snow, D.T. (1965), "A parallel plate model of permeable fractured
media", PhD Thesis, Univ. of California, Berkeley

4,5 Snow, D.T. (1968), "Rock fracture spacings, openings, and porosi-
ties" J.SM&FD ASCE, V94, p73

4,5 Snow, D.T. (1968), "Fracture deformation and change of permeabil-


ity and storage upon change of fluid pressure" Qtly, Colo. School
of Mines, V63,n1

4,5 Snow, D.T. (1970), "The frequency and apertures of fractures in


rock", IJRM&MS, V7,p23

2 Spears, D.D., and Taylor, R.K. (1972), "The influence of weather-


ing on the composition and engineering properties of in-situ coal
measures rocks", IJRM&MS, V9, p729

8 Stacey, T.R. (1969), "Application of the finite element method in


the field of rock mechanics with particular reference to slope stabil-
ity" S. Afr. Mech. Eng., n5, p131

8 Stacey, T.R. (1970), "The stresses surrounding open pit mine


slopes", Proc. Symp. on Planning Open Pit Mines, Johannesburg

8 Stagg, K.G., and Treharne, G. (1970), "Some experiments on the


ultimate bearing capacity of rock= type materials", Proc. 2nd Con g o ISRM
Belgrade, Vl, paper 2-33

4 Stahl, R.L. (1973), "Detection and delineation of faults by sur-


Goodmar-Geologlcal Eng - 30
444 References

face resisti vi ty measurements - Gas Hills region, Wyoming", -U. S. B. Mines


R.I. 7824

4 Stahl, R.L. (1974), "Detection and delineation of faults by


surface resistivity measurements - Schwartzsalder Mine, Jefferson Co,
Colo." U.S.Bureau of Mines R.I. 7975

3 Starkey, J. (1970), "A computer programme to construct spherical


projections", Experimental and Natural Rock Deformation (Springer)

2 Steffen, O.K.H. and Jennings, J.E. (1972), "Definition of design


joints for two dimensional rock slope analysis" Proc. 3rd Congo ISRM,
Denver, V2B, p827

8 Stephansson, O. (1971), "Stability of single openings in horizon-


tally bedded rock", Engineering Geology, V5, p5

7 Stimpson, B. (1970), "Modelling materials for engineering rock


mechanics", IJRM&MS, V7, p77

4,5 Stimpson, B., Metcalfe, R.G., and Walton, G. (1970), "A new
technique for sealing and packing rock and soil samples" Qtly J. Eng
Geol. V3, p127

1 Talobre, J.A. (1967), "La mecanique des roches et ses applica-


tions", second edition (Dunod) - first edition: 1957

8 Tardieu, B., and Pouyet, P. (1974), "Proposition d'un modele de


joint tridimensionnel courbe", Proc. 3rd Congo ISRM, Denver, V2B, p833

2,5 Taylor, R.K. (1973), "Physical and physical-chemical changes


in coal measure rocks", PhD Thesis, Univ. of Durham (U.K.)

2 Terzaghi, K. (1946), "Introduction to tunnel geology" in


Rock Tunneling with Steel Supports, by Proctor and White (Commercial
Shearing and Stamping Co., Youngstown, Ohio)

3 Terzaghi, R.D. (1965), "Sources of error in joint surveys"


Geotechnique, V15, p287

8 Thompson, E. G., and Sayles, F. H. (1972) , "In-situ creep analysis


of room in frozen soil", J.SM&FD, ASCE V98, p899

8 Tincelin, E., Fine, J., and Vouilie, C. T. (1972), "A comparison


of forecasts calculated by the finite element method and the actual
behaviour of mine workings using the small pillar method", Proc. 5th
Int. Strata Control Conf., London

2 Tourenq, C. (1967), "Mise en evidence de la microfissuration des


roches", Compte Rendu des seances de la Societe Geol. de France, V22

2 Tourenq, C., and Denis, A. (1970), "La resistance a la traction


des roches" Lab. des Ponts et Chaussees, Paris, Rapport de Recherche 4
References 445

2 Tourenq, C., and Fourmaintraux , D. ( 1 974 ), "L'indi ce de qualite


des roches, quelques applications" Pro c. 2 nd Con g I AEG, Sao Paulo,
V1, paper 4-20

2 Tourenq, C., Fourmaintraux, D . , and Denis, A. (1971) , "Propaga-


tion des ondes et discontinuites des roche s" , P r o c . I n t . Symp. on Rock
Fracture, Nancy (ISRM)

7 Tufo, M. (1971), "Preliminary invest iga tio n an d c on struction of


a three dimensional geostatical model of t h e Grancarevo Dam", Publ. of
the lnst. of Geot echnics and Foun d. En g., F a cul ty o f Ci vil Engineering,
Sarajevo, Yugoslavi a

4 Trantina, J. A., and Cluff, L .S . (196 3), "NX b o r e h ole camera"


ASTM STP 351

7 Trollope, D.H. (1966 ), "The stability of t r ap e z o idal openings in


rock masses" Rock Mech. and Eng. Geol. V4, p2 32

3,4 Turner, F.J., a n d Weiss, L.E. ( 1963) , "Structur al an a ly sis of


metamorphic tectonite s - Chapter 3 ' Gr aphic treat ment of fab r ic data'"
(McGraw-Hill)

2 Underwood, L.B. (1967), "Clas sific a ti on an d identification of


shales", J SM&FD , ASCE V93, p97

8 Valliappan, S . ( 1968 ) , "Non- l ine a r stress anal y s is of two dimen-


sional problems with special reference to r ock and soil me chanics",
PhD Thesis, Univ. Wales, Swansea (Dept of Civ. Eng.)

4 Van Nostrand, R.G., and Cook, K.L. ( 19 66), "Interpre t a t ion of


resistivity data", U.S.G.S. Prof. Paper 499

2 Vargas, M. (19 5 3), "S ome enginee r ing prope rti es of r e sidual clay
soils occurring in Southern Brazil", Proc. 3rd I nt . Co nf. on Soil Mech.
and Found. En g . , Zu rich, V1, p67

2 Verdier, J. (19 69 a), "Examen u l t rami c roscopi q ue de que lques


roches metamorphiques de sites de ba rrages", 2eme Colloq ue Sur La
Fissuration des Roches, Revue de L'l n dustrie Miner al e , Numero Special
15 July

Verdier, J. (1969b), "Fine microsco pe a nd ul trami cro scope examination


of a few metamorphic rocks from dam s i tes" Eng. Geo l ., V3, p23 3

2 Verdier, J. and Deicha, G. (1 971), " La repartition des inclusions


fluides epigenetiques comme temoin de la microfissur at i o n de s roches"
Proc. lnt Symp on Rock Fracture, Nan cy (l SRM) paper 1-1 0

4 Voloshin ! V., Nixon , D . D., and Timbe r lake , L . L . ( 1 968) , "Orien ted
core, a new technique in engineering geo logy", Bull . A . E .G., V5, p37

2,5 Vutukuri, V.S., Lama, R . D. , and S aluja , S .S. ( 1974) , "Handbook on


446 References

mechanical properties of rocks - - testing techniques and results"


(TransTech Publications)

8 Waddell, G.G. (1971), "Technique of measuring initial deformation


around an opening, analysis of two raise bore tests", U.S.B.Mines
R.I. 7505

8 Waddell, G.G., Crocker, T.J., and Skinner, E.H. (1971), "Tunnel


relaxation method for determining the initial and long term deformation
around an underground opening", Proc. 12th Symp. on Rock Mechanics,
(AIME), p903

1 Wahlstrom, E.E. (1973), "Tunneling in rock-" (Elsevier)

7 Walker, P.E. (1972), "The shearing behaviour of a jointed rock


model", Ground Eng., V5, p24

4 Wallace, R. E. (1950), "Determination of dip and strike by indi-


rect observations in the field and from aerial photographs, J. Geol.,
V58, p269

5 Wallace, G.B., and Olsen, O.J. (1966), "Foundation testing


techniques for arch dams and underground power plants" Testing Tech.
for Rock Mech., ASTM STP 402, p272
2 Walsh, J. (1965a), "The effect of cracks on the compressibility
of rock", J. Geoph. Res. V70, n2

2 Walsh, J. (1965b), "The effect of cracks on the uniaxial elastic


compression of rock", J. Geoph. Res. V70, n2

2 Walsh, J.D. (1974), "Detection of rock defects by fluorescence


photography", CSIRO, Australia, Div. of Applied Geomech., Tech.Memo.12

8 Wang, F.O., Panek, L.A., and Sun, M.C. (1971), "Stabilityanaly-


sis of underground openings using a Coulomb-Navier shear failure
criterion" Trans . SME of AIME, V250, p317

8 Wang, F.D., Sun, M.C., and Ropchan, D.M. (1972), "Computer


program for pit slope stability analysis by the finite element stress
analysis and limiting equilibrium method" U.S.B.Mines R.I. 7685

4 Warren, C.R., Schmidt, D.L., Denny, C.S., and Dale, W.J. (1969),
"A descriptive catalog of selected aerial photographs of geologic
features in areas outside the U.S.", U.S.G.S. Prof. Paper 591

2 Watkins, M.D. (1970), "Terminology for describing the spacing


of discontinuities of rock masses!! Qtly J. Eng. Geol. V3, p193

3 Watson, G . S. (1966), "The statistics of orientation data", J.


Geol., V74, p786

5 Wawersik, W.R. (1975), "Technique and apparatus for strain


measurements on rock in constant confining pressure experiments", Rock
References 447

Mech. (in pres s)

5 Wawersik, W.R., and Brown, W.S. (1973), "Creep fracture of rock"


Report from Uni v. Utah to U.S .B.Mines,(AD-7 38-002)

2 We inbrandt, R.M. , a n d Fatt, I. (1970) , "Scanning electron micro-


scope study of the pore structure of sandstone" , Proc. 11th Symp. on
Rock Mech. (AlME), p629

2 We in e rt, H. H. (1964), "Basic igneous rocks in road foundat ions fI


Bull. National lnst. o f Road Res. V2,pl (S.Afr.)

2 Weinert, H.B. (1965), "Climatic factors af f ect ing t he weather i ng


of igneous rocks" , Agr. Meter. V2, p27

8 West J L . J., and Per ry, R. M. (1969), "Rock mechanics studies for
a high cut at a nuclear power s tation in Penn." Proc. 11th Symp. on
Rock Mech.

4,6 Wickens, E.H., and Barton , N.R. (1971) , f~ pplication of photo-


gramme t ry t o the stability of excavat ed rock slopes l l , Photogrammetric
Record, ( Ph otogrammetry Soc. London) V6, n37

4 Wiebeng a , W.A. and Polak, E.J. (1969), flGreat Lake Nor t h en g in.
Geophys i ca l surveys, Tasmania, 1951-1959", Australia Dept. of National
Developme n t , Bureau o f Min. Resources, Geol, and Geophysics, Bull 101

2 Williams , H., Tur n e r, F. , and Gi l b e rt, C. (1958) , "Petrograp hy" ,


( F reeman )
4 Wil li ams, J.C.C. (1969), "Simple photogramme t ry" (Academic Pr e ss )

8 Wilson, E.L. ( 1972), "Solid SAP , a static analysiS program for


three dimensiona l solid structures", Uni v . of Cal. , Berkeley , Dept . of
Civi l Eng ., Div. of S.E.S.M. Report

4 Wi l son, G. (1967) , "The geometry of cylindrical and conica l f o lds"


Proc. o f the Geologists' Association, V78 , p179

8 Winkel , B.V. Gerstl e, K.H ., and Ko , H.Y. (1972), TlAnalysis of


j

time dependent deformations of openings in sa l t mediaH , lJRM&MS,


V9, p249

5 Withe rs , J.H . (1964), "Sliding res i stance along discontinuiti es


i n rock masses " , PhD Thesis, Univ. of Illinois

6 W1t t k e , W. (1965), "Me thods to analyze the stability of rock


s lopes with and wi thout additional loading " (in German), Rock Mech.
and En g. Geol" Supp. II, p52

6 Wittke, W. (1966), "Berechnungsmoglichkeiten den Standsicherheit


von Bos chunge n in Felsll,(Deutsche Ges ellscha f t fur Erd un Grundbau)

6 Wittke, W. (1967), " Influence of the shear strength of the jOints


448 References

on the design of prestressed anchors to stabilize a rock slope", Proc.


Geotech. Conf., Oslo, paper 4-11, p311

8 Wittke, W., Rissler, P., and Semprich, S. (1972), "Three dimen-


sional and turbulent flow through fissured rock according to discon-
tinuous and continuous models" (in German), Proc. Symp. on Percolation
Through Fissured Rock, Stuttgart (ISRM)

8 Wi ttke, W., Rodatz, W., and Wallner, M. (1972), "Three dimension-


al calculation of the stability of caverns, slopes, and foundations in
anisotropic, jointed rock, by means of the finite element method",
Deutsche Geotechnik, Vl, n1

8 Wright, F.D. (1972), "Arching action in cracked roof beams",


Proc. 5th Int. Strata Control Conf., London

8 Wylie, C.R. Jr. (1960), "Advanced engineering mathematics",


(McGraw-Hill)

8 Yoshida, M., and Yoshimura, K. (1970), "Deformation of rock mass


and stress in concrete lining around the machine hall of Kisenyama
underground power plan t" Proc. 2nd Cong. ISRM, Belgrade, V2, paper 4-29

8 Zienkiewicz, O.C. (1967), "The finite element method in struc-


tural and continuum mechanics" (McGraw-Hill)

8 Zienkiewicz, O.C. (1970), "Three dimensional finite element


analysis of the Tachien arch dam", Water Power, V22, p173

8 Zienkiewicz, O.C. (1971), "The finite element method in engineer-


ing science" (McGraw-Hill)
8 Zienkiewicz, O.C., Best, B., Dullage, C., and Stagg, K.G. (1970),
"Analysis of non-linear problems in rock mechanics with particular
reference to jointed rock systems", Proc. 2nd Congo ISRM, Belgrade,
V3, paper 8-14

8 Zienkiewicz, O.C., and Cheung, Y,K. (1966a), "Application of


the finite elemen t method to problems of rock mechanics", Proc. 1st
Congo ISRM, Lisbon, Vl, p661
8 Zienkiewicz, O.C., and Cheung, Y.K. (1966b), "Stress in aniso-
tropic media with particular reference to problems of rock mechanics -
- finite element analysis" J. of Strain Analysis, Vl, p172

5 Zienkiewicz, O.C., and Stagg, K.G. (1966), "The cable method of


in-situ testing", Proc. 1st Congo ISRM, Lisbon, Vl, p667

8 Zienkiewicz, O.C., Watson, M., and King, I.P. (1968), "A numeric-
al method of viscoelastic stress analysis", IJRM&MS, VI0, p807

8 Zienkiewicz, O.C., Valliappan, S., and King, I.P. (1968), "Stress


analysis of rock as a no tension material" Geotechnique, V18, p56
References 449

4 Zimmer, P.W. (1 96 3), "Orientation of s mal l diameter dr i ll c ore",


Econ omi c Geology , V58, p1313

ABBREVI AT IONS

A.E.G. Association of Eng i n e e ring Geol ogi sts


A.I.M.E. American Instit ute of Mi ning and Metall u rgy
A.S.C.E. American Societ y of Ci vi l Engine e rs
A.S.M.E . Ame ric a n Socie ty of Me chanic a l Enginee r s
A.S.T.M. American Society f or Te stin g and Mate r ials
Can. Cana dian
C.S. loR.O. Ce n tral Scientific a nd I ndustrial Resea rch
Organizati on ( Aus t ra li a )
G.S.A. Geological Society o f Ameri ca
I.A.E.G. Internation al Associat i on of En gineeri ng
Geologist s
IJRMSu\1 S Internationa l Journ al of Rock Mechan i c s a nd
Mining Scie n c e
I.M.M . Institution of Min in g and Me tallurgy (Lo n d o n )
J. Journal
R. I. Report of I nve stigation
S.E.S . M. Structural Engineer in g and Struct u ral Me chan ics
(University of Cali f ornia, Dept. of C . E .)
S.M.E. Society of M~ ning Eng ineers (A I ME)
SM&FD Soil Mechanics an d Foun d at ions Divisi on (ASCE )
S. T .P. Special Technical Publ icat i o n
U.S.B.M i nes Un ited States Bu r eau of Mines, U .S . Dep artmen t
of the Int erior
U.S.G.S. Un i ted States Geolog i cal S urvey, U.S . De part ~

ment of the I n t eri or


...
subject -ndex

acceleration, 242 - 247

accessory minerals, 19, 35

active and passive blocks, 261-270

aerial photo interpretation, 102-104

aerial photos - scale, 102

- sources, 109, 110

amphibole, 35

anhydrite, 24

anisotropic friction, 238, 239

anisotropy, 15, 18, 200-208, 312-316

aperture, 44

apparent dip, 101

area of a face, 227

area of contact correction in shear tests, 163

asperities, a78, 180, 185-191

argillite, 42

augering, 129, 130


452 Subject Index

"back calculations", 169, 170

banding, 15, 40

basalt, 15, 33, 37, 48, 184

base friction models, 277-284

bedding, 15, 18, 40, 42, 44, 45, 103, 254

blasting, 8, 9

"blind zone" of an outcrop or a borehole, 88

bore-hole cameras and television, 153, 154

bore-holes for exploration, 127-142

Bray's diagram, 201-205

Brazilian test, 35, 36

Brunton compass, 95, 96

cable reinforcement, 168, 254, 297-298

calcite, 35, 46, 183

case histories - back calculation, 169, 170

cement, 18, 46

centroid of a wedge, 228

chalk, 19, 48

Chambishi mine, 9

chert, 19, 33

Chivor dam, 8, 44, 45

chlorite, 19, 46, 183

Christensen orienting core barrel, 150, 151

claystone, 29

cleavage, 40

coal, 19, 129

cohesion, 170, 243, 244, 251, 252


Subject Index 453

comminuted rock, 44

common excavation, 23

compound surfaces, 181-182, 192-193

"consolidated" ro ck, 2 0

consti tu tive relations - for joint c ompression, 1 7 1

- for jOint shear ing, 174

contact area, ac t ual , 170

contacts, 40

continuous rock masses, 1, 2, 47-49

contouring poles to joints, 83

core barrels , 1 33- 1 39

core recovery, 133, 137

core sizes, 134-135

Craelius core orienter, 149-150

cross-hole shooting , 12 3

dams, 5, 283, 2 8 6, 287

dam abutments - stability analysis, 2 5 5- 2 6 1

"daylighting" cond i tion, 21 2- 2.15, 255

decomposed rock, 20 , 2 1

diamond drilli n g, 1 30-133

dilatancy - duplication in finite element an alysi s , 34 7-348

- effect of boundary conditions, 177, 1 86 , 195

- importance, 2, 47

- in models, 2 90

- in rotational sl idi ng , 19 3-196 , 2 00

- measurement, 166

- relat ion to roughness, 176-183


4~ Subject Index

- variation with normal stress, 186-191

dilatometers, 157

dip and strike - explanation, 96-97

- measurement in aerial photos, 104-109

dipmeter, 154

direct shear tests - in-situ, 168-169

- laboratory, 165-168
discontinuities - aperture, 44, 153

behavior in compression, 173

behavior in shear, 173-176

- filling material, 44

- in granitic rocks, 103

- spacing, 43-44, 49, 51-52

- roughness and waviness, 41-42, 52, 55

- sampling methods, 158-160

- simulation in models, 278-280, 290

- single versus multiple features, 211

- statistical treatment of orientations, 83-86

- types, 40

- stresses in, 49, 53, 271-275

direct stiffness method, 301

discontinuous rock masses, 1-2, 15, 47-49

- equivalent anisotropic continuum, 315

distorted models, 277, 289

dolomite, 20, 129, 184

Don Pedro d am, 44-45

drainage, 10
Subject Index 455

dri ll i ng - core orj.e ntation, 1 42-155

- cos t , 127

- f or joint sampling, 1 58-1 59

- methods, 129- 1 33

- p robability of intersecting a t arget, 128-129

durabil i ty , 27-30

dr i l1ability, 16, 50 , 1 29-13 1

e a rt hquake f orces. 243-247

e q u a l area proj ection, 65

e l e ctrical r e si s tivity, 1 25-126

el ement s t i ffne s s mat rix - constan t s train triangle , 307-308

- join t e leme nt, 327

e xcavat ion - c las sii i cati on of mater i a l, 23

e xcavations - surface, 5

- u n derground openings, 5, 7, 285, 290-293 , 365-367

extension fractures, 171

f ac t or o f safe ty - de f inition , 24 1

- in t he int e rs e ction mode, 250-255

- relat ion t o probabili t y of safety , 243, 252

f aul t s J 4, 4 4-45, 48-49 , 103, 125-126

fe l dspar, 19, 22, 25 , 26 , 35 , 183

f ed l spat h o ida1 minerals , 19

"f i l l ed" joints , 1 74-175

f ini t e eleme nt analys i s - a c celerations , 3 53

- band width , 370

- conve rgence, 355-358. 370 372


Subject Index

- displacement boundary conditions, 305-306, 372

- gravity loading, 353, 363-365

- incremental loading, 350, 367-368

- joint closing, 341-342, 354-357

- joint dilatancy, 347-348

- joint element stiffness, 320-329

- jOint opening, 336-340, 357-358

- joint shearing, 342-347, 358-362

- load transfer method, 334-335, 370

- modified joint elements, 335-336

relevance to geological engineering, 300

- rock bolts, 354

- solution of equations, 370

- structural stiffness matrix, 301, 304, 330-331

- variable stiffness method, 334-335, 370

- water forces, 349

Fisher's distribution, 86

filling material of joints, 44, 52, 55

fissures, 16-18, 20, 30-39

- contrasted with discontinuities, 40

- contrasted with pores, 30-32

- defined by spacing, 44

- microscopic study, 31, 33

- porosity, 33-36

flexural cracking, 279, 285-287

focal length of a lens, 108, 113, 117

foliation, 15, 18, 40


Subject Index 457

following, versus non-following, forces, 280

form ation s, 47

friction, 183- 187

friction circle concept, 23 7- 240

GeoNor swelling apparatus, 29

geological mapping, 93-96

geological maps - scales, 93-94

- sources of existing maps, 93

- symbols, 92, 94, 97

Geological Society of London, Enginee r ing Grou p, 93

geophysical methods, 121-127

gneiss, 39, 47, 142, 184

gnomonic projectio n , 114-117

"Goodman" jack, 157

gouge, 46, 184, 192

Granc arevo dam, 294-29 9

granite, 15, 18, 22-26, 31, 33, 3 7, 39, 4 2 , 47 -48, 103, 1 23, 1 84

graphi te, 5, 46

gravity problems, 277-284, 352-35 3

gravi~y surveys, 1 26

Griffith theo r y, 32

g r outing, 5, 10

gypsum, 19- 2 0, 24, 46

hardness of joint surfaces, 43, 191

hematite, 127

hydrothermal alteration, 22
458 Subject Index

imbricated block structure, 195, 290, 296

integral sampling method, 151-153

intermediate principal stress - effect on strength, 208

internal and external forces - 272-273, 303

International Society for Rock Mechanics (ISRM), 12

intersection line of two planes, 75-76, 212-214

intersection mode of sliding, 247-261, 265-270

joint blocks, 41, 43, 210-211

joint surveys, 83-88

joints, 40, 42, 44

- artificial, 159-160, 171

- columnar, 2, 43-45

- deformation measurements, 161-163

- extent, 41, 102-104

- filled, 174-175, 191-192


_. healed, 44

- incipient, 174

- initial stresses in, 329-330

- normal stiffness, 338

- duplication in models, 278-280, 290, 296

- samples, 158-160

- shear stiffness, 174

- sheeting, 44-45

- stresses in, 49, 322-325

Karst, 24

kinematic models, 277-284


Subject Index 459

kinematic tests for sliding under gravity load, 212-217

kink band formation, 196-199

Kukuan dam, 5

Lambert net, 65

landslides, 4-5, 48-49

latite, 44, 45

Libby dam, 44-45, 254-255

limestone, 20, 24, 37, 39, 42, 48-49, 129, 184, 295

limit equilibrium analysis, 240-270

logging drill holes and core, 139-142

logging excavations and test pits, 118, 278

loosened rock, 48-49, 284

Los Angeles abrasion test, 29

magnetic methods, 126-127

magnetite, 127

Malpasset dam, 5, 39, 255

marble, 24, 33, 47

maximum joint closure ( V ) , 171-173, 289


mc
mica, 18-19, 35, 46, 183

minerals, 19, 35

mining, 4

models, 10, 277-299

- materials, 278-280, 289-290, 296

- scaling laws, 288

- two versus three dimensions, 293-294

modulus of elasticity, 34-36, 289, 312-316


Goodman-Geological Eng -31
460 Subject Index

Mohr's circle diagram, 161, 207, 191

montmorillonite, 19, 183, 192

multiple block slides, 261-270

nodal pOlnts, 301-302

normal deformation of joints, 170-173, 336-342

Norwegian Geotechnical Institute (NGI), 29, 51

opal, 19

oriented core, 142-155

- using stereographic projection, 144-147


Oroville dam, 7

outcrop patterns of planar features, 96-104

peak shear strength, 174-175, 183-200

pegmatite, 33

petrology, 16-19

photo base, 106

pitch angle, 76

point load test, 30, 50-51

Poisson's ratio, 34-35, 39, 312-316

pore pressure, 167, 192, 199-200

porosity, 126

principal point of a photo, 104-105, 112-113

probability of safety, 239-240, 242-243

prototype, 277

pump-in pressure tests, 155, 157

pyrite, 19

pyroxene, 35
Subject Index 461

pyrrhotite, 19, 127

quartz, 19, 22, 25-26, 35, 46-47, 183

q uartzite, 2, 5, 17, 33, 39, 47-48

radar, 110

radial permeability test, 30, 36, 38-39

remote sensing, 109-111

res idua l shear strength, 174-175, 192, 198, 346


resistivity logging, 154-155, 157

resistivity method, 125-126

rigid block a n alysis, 368

rippability, 23, 26, 122

ripple marks, 42

rock bolts, 10, 241, 251, 254, 354

rock classification - for different purposes, 50

- for excavation, 22, 23

- geological, 16-19

- "solid" rock, 20-21

rock description, 56-57

rock mass, 15-16

rotation analysis, using the stereonet, 231-237

rotational sliding, 192-199

roughness, 41-42, 52, 55, 178-183

- measurement by s catter diagram of poles, 181-183

- relation to dilatancy, 183-185, 190

RQD, 51-52

Ruedi dam, 44-45


462 Subject Index

samples of joints for lab testing, 158-160

sandstone, 1, 15, 17, 31, 48, 103, 129, 184

saprolyte, 23-24, 123

scale effect, 32, 175-176, 289

schist, 17, 33, 39, 142

schistocity, 40, 44-45

Schmidt hammer, 30, 43, 191

seams, 174

seismic refraction method, 122-125

seismic velocity, 122-125, 157

Seisviewer, 154-155

serpentine, serpentinite, 19, 46, 129, 183

shale, 1, 15, 18-19, 24, 29, 33, 129

shear strength - peak and residual defined, 174-175

sheared zones, 40

sheet joints, 44

shot drilling, 130-131

sign convention - cross product on the stereonet, 218

- normals to faces of a wedge, 255-256

- positive direction of a line of intersection, 215, 255-256

- shear and normal stresses, 272-274

- tension positive in finite element program, 307

similitude, 281, 288

slake durability test, 28-29

slate, 5, 47, 183

soil-like rock, 19-22, 25

soils, 2, 19-22
Subject Index 463

sp acing of d is c o nt inu ities , 4 3- 44, 49, 51-52, 10 1

s pillways , 8

squeezing g round , 46

s t and-up-t ime, 46-5 5

s tate o f s t res s, 27 1-2 76

s tat i st ical t reatmen t o f joint a tt i t u d es, 83-86, 1 57

sta t is t i cs of s earch, 1 27- 1 28

s tereographi c proj e c tion - add it i on and subt r ac tion of v ec tors, 218-222

- additio n of wate r forces , 22 1-2 22

- a nalysi s of rotation, 231 -23 7

- angles be t ween lines J 73-75

- conf o rmal prope rty - 60


- contour plots f or p o l e d iagrams, 8 3-86

- cross produ c t of ve c t ors, 217-2 1 9

- decompositi on of a v ector , 222-223

- to define the g eometr ic pr op e rt i e s of a we d ge , 22 4- 228

- dot pr o duct , 218-219

- f ocus , up pe r and l ower hemispher e projection s , 59

- friction, 237-2 4 0

- intersec t i on li ne of two p l anes , 7 5 -7 6

- lines in space , 59- 61 , 69-7 1

- opposi t e to a line , 6 9 -71

- pierc i ng p o int of a f o rc e on a pl ane , 228-231

- p lan e and it s no r mal ! 5 9- 61, 71-73

- Ilp rimitive c i r cle"

- rotation of a line , 80-83

- roughne s s est imation , 1 81- 18 3


464 Subject Index

- shear and normal stresses on a discontinuity plane, 271-276

- small circles, 61-62, 78-80

- stereonets, 62-68

- test for intersection slides under gravity, 212-214

- test for plane slides under gravity, 214-216

- upper and lower hemispheres, 59, 71

stiffness - meaning of the term, 301-302

Stockton dam, 9

strength reduction to account for jointing, ~58

stress dependency of physical properties, 32

stress-strain relations, 312-316

stresses - initial, 8, 10, 15, 18, 49, 270, 317-318, 329-330, 367-368

strike, 96-97

surveying for geological mapping, 95

swellability, 29

swelling clay, 46

swelling rocks, 20-22, 24

symbols for geological maps, 97

Symposia on rock mechanics, 12

talc, 19, 46

tapered blocks, 216-217

tensile strength, 35-36

Televiewer, 154-155

tension in jointed rock, 171, 202-204

terrestrial photogrammetry - principles, 112-117

- use in logging geological features, 117-121


Subject Index 465

text u re, 16-18, 33

three pOint problems, 99

trachyte, 175

transformation of stresses, 369-374

triaxial tests, 160-165

triple tube core barrel, 137-138

Tullock dam, 44-45

tunneling - rock classifications for, 50-55

two block slide analysis, 261-270

underground openings, 5, 7, 285, 290-293, 365-367

Vajont reservoir, 5

wash boring, 129-130

water - effect on friction, 183, 184

water forces - calculation of effect on stability, 244, 260-26 1

- in finite element analysis, 349-352

weatherability, 18, 27-30

weathering - degree of, 15, 25-27, 191

- products of, 22-25

- profiles, 27

- slaking, 1, 28-29

wedge failure - examples, 5, 254-2 5 5

- kinematic tests, 21 2-214

- stability analysis with one free face, 254- 26 1

- stability analysis with two free faces, 247-2 54

wire sawing joint samples, 159


466 Subject Index

Wulff new, 62-63

yield acceleration, 243-247

zeolite, 19
author index

Aastrup, A. , 19 Bieniawski , Z. ,' 50, 54 , 55

Archambault, G., 189, 190, 193, Bjerrum, L., 29

l_95-199, 346, 373 Blindheim, O. , 50

Arnold, K. J., 83, 84 Blyth , F. G. H. , 9l , 96

Ashby , J., 196 Bolstad, D. D. ,

Atkinson, K. B. , 112 Bowden, F. P,, LTL


Badgeley, P. C. , 91 Boyd , J., 181, 182

Baltosser, R. , L54, 155 Brace, 32


Barr, D. J. , tIO Bradley, D., 33

Barroso, M. , 153 Brawner, C. , 169

Barton, N., 50-54, it76, 180, 181, Bray, J., 8, L2, 159, 166, 169, L70 ,
r9o, 191, 196, 289, 29O 193, 2OL-2O5, 2I3, 2L4, 234, 28L
BerentS , H. P. , l-37 Brekke, T. , 29, 46, l4O, l-75
Bergh-Christensen , J., 50 Broch, E., 30

Bernaix, J., 5, 30, 32, 36, 38, Brune, , 24


G.
39, 166, 198 Burman, B. C., 166, 37O

Berry, L. , 23, 27 Burton, A. , 43


468 Author lndex

Byerlee, J. D. , 160 , 183 Eaton, G. P., L26


Byrne, R. J., 368 Ege, J. , 26, 43, 50
Camargo, F. , L69 Eureni-us,J., 191

Cassines, R. , 109 Evdokimov, F. D. , 166

Chenevert, , 2O5
M. Everling, 2gg
Clough, R. W., 308 Fagerstrorn, H. , 191
Cluff, L. S., 111, 154 Fairhurst , C. , 189, 37 3

Coates, D. , L2, 49, 5O Fanshawe, H. G., L23


Cook, K. L., L22 Fatt, I. , 31
Cook, N. G. W. , 13 Fecker, E., 181, 182
Coulson , J. fl. , 173, 183, 184 Fisher, R. A. , 83, 84, 86
Cumming, J. D., 145 Fookes, P. , 2O, 27 , 28
Cundall, P. , 196, 368 Fourmaintraux, D. , 34-36
Dearman, W. , 2O, 28, 92 Franklin, J, 2O , 28, 3O, 49 , 50, 51
,
DeChambrler, P., 155 Fumagalti, E., 288
Deere, D. IJ. , 23 , 27 , 30, 49 , 50, Gale, J. , 349
51, 183, 193, 289 Gamble, J.,29
Deicba, G., 33 Gatlin, C. , 2O5
Del Vesco, E., 166 Geyer, R. L,, L57
Denis, A. , 34-gT Gilbert, C. , 16
Dobrin, M. 8., L22 Grant, F. S., LZL
Donath, F. A., 2O5 Griffith, J., 27
Donn, W. L. , 91 Grif f iths, D. H. , L24
Dorn, W. S., 334, 335 Habib, P. , 32, 38
Dubois, J., 4L, g25, 335, 365 Hamrol, A. , 27
Duncan, J. M., 30, 315 Han, C. Y., 352
Duvall, W. , 13, 50 Handin, J., 160, 200
Author lndex 469

Haverland, M. L. , 168 KJ-tihara, Y. , L97


Hayashi, M. L97 Kobayashi, S. , 187
Heck, W. J. 160 Krsmanovic, D., L66, I73, 2gO,

Heiland, C. A., J-22 294-299


Hend.ron, A. J., Jr., 246 Kruse, G. , 50

Heuze, F. , I51, 160, 2L6, 2L7 Krynine, D. , 13


Hoek, E. , 8, ,
, 159 , 166 ,
L2 32 Kutter, H. K., L67
169, L7O , L73 , 196, 2L3 , 2L4 , Ladanyi, B., 189, 190, 191, 193,
280 195-199, 346 , 354, 373
Hoffman, H. , 50 Lajtai, E. Z, , I84
Horn, H. N., 183 Lane, K. S., 160
Horswill, P. , 27 Lang, T. A., lI9, 365
Howard, T., 4 Langof, 2., 166 , I73, 294-299
Iida, R., 50, J-87 Lauffer, H. , 4O
Iliev, I. , 25, 26 Leet, L. , 4Q

Irvlng, E. , 84 Lien, , 50-54


R.
Jack, II. , 278 Linkwitz, K. , 119
Jacquet, P., 33 Locher, H. G. , 166
Jaeger, J., 13, 160, L75, I83, Lokin, P., 9O
r88, 189, 193, 2O5, 2O8 Lombardi, G. , 160
Jakosky, J. J., L22 Londe, P. , 22O, 23L, 234, 248, 25O,
Jeran, P. , 83 255, 256
John, K. , 50 , L75, 2O5, 248, 252 Lopes , J. B. , 168

Judd, W. , 13 Lugeon, , 155


M.

Judson, S. , 40 Lumb, P. , 23, 25, 27

Kiersch, ,
G. 26 Lunde, J., 50-54
King, R. F., L24 McCracken, D. D., 334, 335
470 Author lndex

McGregor, K. , 130 Patton, F. , 23, 27 , 4L, I8I, 185,


McMahon, B. K., 83 L87

Mafrtab, M. A. , 83, 84 , 86, 87 , 96 Pender, M. , 27

Maini, Y. N. T., 155 Pentz, D. , J-73

Markland, J. , zlg Phillips, F. , L46, 155


Martna, J., 19 Pincus, H. J.r 83

Mashey, J. , 83 Polak, E. J., L22, L23, L25, L27


Matthews, K., L44 Price, N. , 4Q
Mei-ssner, R. , L24 Ragan, D. M., 91
Mencarelli, E. , 33 Raleigb, C. B., 160
Moffitt, F. H., 105, 113, 114 Ramsay, J .., 196, 25? , 253
Morlierr,P. , 33 Ray, R. G. , 108
Mi1ler, R., 30, 49, 5O Rengers, N., 119, 166, 178, L79,
Moye, D., 139 181, 182

Mueller, W. A. G. , 119 Richards, L. P.

Miiller, L. , 6, 50, 2O5 Rocha, M. , 151 , L52, 153 , L5?

Myung, J. , L54, 155 , L57 Rodatz, W. , 351

Nascimento, U. , 29 , 183 , 193 , 194 , Rosengren, K., 143, L49, 160, L62,
199 163, , L73 , 183 , 208
L64

Noorishad, J. , 38, 34g Ross-Brown, D. M. , 108, LLz, 119,


Obert , L. , 13, 50 181

Ohnishi, Y., 160, 166, L67, 2OO Rowe, P. W., 189


Onodera, T., 50 Ruiz, , L69
M.

Panet, M. , 31 Rummel, F., 160

Parsons, R. Ruxton, B., 23, 27

Paterson, M. S. , 160, 196 Sabarly , F. , 155, 156


Author Index 471

S aho re s , J. , 2 4 Te r zagh i , R. , 50, 86- 88

S t. Joh n , C. , 369, 39 4 Te ixe ira , H. , 183 , 193 , 194

Sal1st r om, S . , 1 9 To u r e n q , C. , 3 1, 3 3-37

S apegi n , D. , 1 6 6 T ran t ina , J. , 154

Savage , J. F. , 1 19 Treashe r, R . , 26

S c hnei d e r , B. , 31, 3 3 Tr o 1 1ope , D. H. , 28 0

Sco tt, R. , 1 3 Tuf a , 294-2 99

Seed, H. B . , 246, 247 Turne r , F., 16 , 83

Selmer-Ol sen, R., 5 0 Underwood , L., 50

Ser afim , J. L. , 27, 168 Van , T. K ., 1 5 7

S ha n l ey, R. J. , 86 Van Nostran d, R. G. , 1 22

Shimer, J. A. , 9 1 Var gas , M" 23

Simp son , B ., 91 Verdi e r, J., 3 1, 33

S l ebir, E. J ., 1 68 Vigier, G., 220, 23 1

S l e mmo n s , D. B. , I I I Vo losh in, V., 1 50

S l ichter, L., 1 2 8 Vo r me r i ng e r, R., 22 0 , 23 1

Snow , D. T ., 1 5 6 Wal lace, R. E., 109 , 114

Spe ars, D., 27 Wa lsh , J' i 32, 33

S tagg, K. G., 1 68 Wa tki ns, M. D. , 44, 12 3

Stah l , R. L., 125 Waw e r s ik, W. A' 1 1 63

Stearns, D. W" 160, 200 We i nbr a ndt , 31

S t i mps on, B., 289 Wei n e rt, H., 20

Tab 0 r, D. , 1 71 We i s s, L., 8 3 , 19 6

Ta1o bre, J. A" 88-90 West, G. F., 1 21

Taylor, R. L., 27 , 4 6 , 1 7 5 , 231 Wi e ben g a, W. A. , 1 27

Terzagh i , K. , 2 0 , 21 Williams , H. , 1 6
472 A u thor Index

Williams, J. C. C., 112, 119 Wylie, C. R., Jr., 301

Wilson, E. L., 370 Zeltinger, J., 278


Wittke, W., 213, 231, 234, 248, Zienkiewicz, O. C., 168, 308, 351
349 Zimmer, P. W., 154

You might also like