FOR MONITORING
FIELD PERFORMANCE
GEOTECHNICAL INSTRUMENTATION
FOR MONITORING
FIELD PERFORMANCE
John Dunnicliff
Geotechnical Instrumentation Consultant
Lexington, Massachusetts
Gordon E. Green
Geotechnical Engineer
Seattle, Washington
[I]
WILEY
A Wiley-Interscience Publication
John Wiley & Sons
New York / Chichester / Brisbane / Toronto / Singapore
Copyright © 1988 by John Wiley & Sons, Inc.
All rights reserved. Published simultaneously in Canada.
Reproduction or translation of any part of this work
beyond that permitted by Section 107 or 108 of the
1976 United States Copyright Act without the permission
of the copyright owner is unlawful. Requests for
permission or further information should be addressed to
the Permissions Department, John Wiley & Sons, Inc.
Every geotechnical design is to some extent hypo- desire to save money, can be more than false econ-
thetical, and every construction job involving earth omy: it can even be dangerous.
or rock runs the risk of encountering surprises. Every instrument installed on a project should be
These circumstances are the inevitable result of selected and placed to assist in answering a specific
working with materials created by nature, often be- question. Following this simple rule is the key to
fore the advent of human beings, by processes sel- successful field instrumentation. Unfortunately, it
dom resulting in uniform conditions. The inability is easier to install instruments, collect the readings,
of exploratory procedures to detect in advance all and then wonder if there are any questions to which
the possibly significant properties and conditions of the results might provide an answer. Instrumenta-
natural materials requires the designer to make as- tion is currently in vogue. Some design agencies
sumptions that may be at variance with reality and and many regulatory bodies mandate instrumenta-
the constructor to choose equipment and construc- tion whether the results might be useful or not. It is
tion procedures without full knowledge of what a widely held dogma, for instance, that every earth
might be encountered. dam should be instrumented, in the hope that some
Field observations, including quantitative mea- unsuspected defect will reveal itself in the observa-
surements obtained by field instrumentation, pro- tions and give warning of an impending failure. Part
vide the means by which the geotechnical engineer, of the criticism directed at Teton Dam following its
in spite of these inherent limitations, can design a failure was paucity of instrumentation. Yet, it is
project to be safe and efficient, and the constructor extremely doubtful that any instrumental observa-
can execute the work with safety and economy. tions could have given timely warning of the partic-
Thus, field instrumentation is vital to the practice of ular failure that occurred. Instruments cannot cure
geotechnics, in contrast to the practice of most defective designs, nor can they indicate signs of im-
other branches of engineering in which people have pending deterioration or failure unless, fortuitously,
greater control over the materials with which they they happen to be of the right type and in the right
deal. For this reason geotechnical engineers, unlike place.
their colleagues in other fields, must have more than The engineer should bring the best knowledge
casual knowledge of instrumentation: to them it is a and judgment to bear on every geotechnical prob-
working tool, not merely one of the components of lem that arises and should analyze the quality of the
research. information on which a design is based. The en-
Notwithstanding its vital role, instrumentation is gineer should judge not only the way the design will
not an end in itself. It cannot guarantee good design function if the information is essentially correct, but
or trouble-free construction. The wrong instru- how the gaps or shortcomings might influence the
ments in the wrong places provide information that performance of the project. Then, and only then,
may at best be confusing and at worst divert atten- can specific items be identified that will reveal
tion from telltale signs of trouble. Too much in- whether the project is performing in accordance
strumentation is wasteful and may disillusion those with design assumptions or, if not, in what signifi-
who pay the bills, while too little, arising from a cant way the performance differs. Then the critical
vii
viii FOREWORD
questions can be framed-the answers to which will sure that dependable data can be obtained through-
fill the gaps or correct the errors in the original de- out the period when the observations are needed.
sign assumptions-and the engineer can determine Usually, the most dependable devices are the sim-
what instruments, at what locations, can answer plest.
those questions. If one can with sufficient accuracy make a direct
Of course, not all instruments are installed to visual observation with a graduated scale, then a
monitor the safety of a structure or construction micrometer should not be used. If one can use a
operation or to confirm design assumptions. Some micrometer, a mechanical strain gage should not be
are used to determine initial or background condi- used. If one can use a mechanical strain gage, an
tions. Observations of groundwater prior to con- electrical one should not be used. Mechanical in-
struction, or in situ stresses in rock masses, or of struments are to be preferred to electrical devices
elevations of structures before the start of adjacent and simple electrical devices depending on simple
construction are examples. Certain types of con- circuits are to be preferred to more complex elec-
struction, such as the installation of tiebacks, are tronic equipment. That is, where a choice exists,
inherently dependent on instrumentation. Further- the simpler equipment is likely to have the best
more, advancements in the state of our knowledge chance for success.
require large-scale or full-scale observations of an Nevertheless, simple instruments are sometimes
extent and complexity far beyond the requirements inappropriate and more complex ones must be
of the practicing engineer. Yet, in all these applica- used. An open standpipe may be the simplest de-
tions, it is equally true that every instrument should vice for observing a piezometric level, but the point
be selected and located to assist in answering a at which the pore pressure needs to be measured
specific question. may be located where direct access is impossible. A
Instrumentation needs to be kept in perspective. more sophisticated arrangement is then necessary.
It is one part of the broader activities of observation If one wishes to determine the state of stress in a
and surveillance. Trained people, using the best of mass of rock, there is no choice but to install
all instruments, the human eye, can often provide sophisticated equipment to make measurements at
all the information necessary and are always an es- a considerable distance from the position of the ob-
sential part of the field observations on any project. server, and the strains likely to be observed will be
Even when instruments are used because the neces- too small to detect by any mechanical device. Thus,
sary quantities are too small to be observed by eye, nothing but a sophisticated system will serve.
or the events are taking place out of the reach of a Not all sophisticated systems are equally reli-
human being, the findings must be related to other able. Equipment that has an excellent record of per-
activities. Without good records of the progress of formance can be rendered unreliable if a single
excavation and of details of the excavation and essential but apparently minor requirement is over-
bracing procedures, for example, the results of looked during the installation. The best of instruc-
measurements of deformations or earth pressure tion manuals cannot provide for every field condi-
associated with a braced cut become almost mean- tion that may affect the results. Therefore, even
ingless. Among the most valuable uses of instru- slavish attention to instructions cannot guarantee
mentation are empirical correlations between con- success. The installer must have a background in
struction procedures and deformations or pres- the fundamentals of geotechnics as well as knowl-
sures, correlations that can be used immediately to edge of the intricacies of the device being installed.
improve the procedures so as to reduce the move- Sometimes the installer must consciously depart
ments or pressures. Thus, highly sophisticated, from the installation manual.
fully automated installations for obtaining and pre- The installer must also want desperately to do
senting data, sometimes held in high favor by those the job well and must often work under difficult and
intrigued with gadgetry, may fail to serve a useful unpleasant conditions, trying to do precision work
purpose because the simple visual observations of while surrounded by workers whose teamwork or
what may be affecting the readings are overlooked. operation of equipment is being interrupted, or
Not only is instrumentation not an end in itself, working the graveyard shift in an attempt to reduce
but neither is sophistication or automation. The two such interruptions. Dedication of this sort is the
prime requirements are sensitivity sufficient to pro- price of success, and it is rarely found at the price
vide the necessary information and reliability to en- tendered by the lowest bidder. Moreover, the in-
FOREWORD ix
staller can hardly be motivated to be dedicated to the struts for bracing open cuts were measured, it
the task of installing instruments of inferior quality was observed that struts at the same elevation in the
that are likely to fail prematurely or to produce same cut carried widely different loads. To some
questionable data. Rugged, reliable instruments are extent the difference was the result of slight differ-
not necessarily expensive, but lowest cost of the ences in soil properties. To a much greater extent,
hardware is rarely a valid reason for its choice. No however, the differences were associated with con-
arrangement for a program of instrumentation is a struction procedures. When a strut was placed as
candidate for success if it sets cost above quality of soon as possible, it always carried a load substan-
instruments or fee above experience and dedication tially greater than the load in another strut at the
of the installer. same elevation but not placed until excavation had
Instruments are discontinuities, nonrepresenta- advanced considerably beyond its future location.
tive objects introduced into soil/rock structure sys- If the strut loads had been measured on only a few
tems. Their presence or the flows or displacements struts, for example, along a single vertical line, the
required to generate an observation alter the very influence of the construction procedure on strut
quantities they are intended to measure. The alter- loads could not have been detected, and entirely
ation may be significant or negligible; its extent de- erroneous conclusions might have been drawn
pends on the nature of the phenomenon being ob- about the magnitude of the earth pressures resisted
served, on the design of the instrument, and on the by the bracing. Since the variability of the quan-
operations required for installation. The engineer tities measured by geotechnical instrumentation de-
who embarks on a program of field instrumentation pends not only on the kind of measurement itself-
needs to understand the fundamental physics and whether it be pore pressure, displacement, or load
mechanics involved and how the various available in a structural member-but also on the geology
instruments will perform under the conditions to and on details of the construction procedures, the
which they will be subjected. In addition, the en- design of a system of measurements requires ma-
gineer needs to know whether corrections can be ture judgment based on experience and understand-
made by calibration or by theoretical calculations, ing of the geotechnical problems at hand.
or whether under the circumstances no valid result Use of field1nstrumentation therefore requires a
is possible. Perhaps the classic examples of the lat- thorough grounding in geotechnical principles, a de-
ter eventuality were the attempts in the early 1920s tailed conception of the variations that may be ex-
to measure earth pressures against braced cuts by pected in the natural or artificial deposit in which
observing deflections of the wales between struts, a the observations are to be made, a realistic notion
procedure made futile because of the then unsus- of the construction procedures likely to be fol-
pected phenomenon of arching. The same phenom- lowed, a thorough knowledge ofthe capabilities and
enon affects to greater or lesser degree the results of shortcomings ofthe instruments themselves, and an
all earth-pressure cells. appreciation of the practical problems of installa-
Finally, there must be enough instruments not tion. It also requires a clear perception of the way in
only to allow for the inevitable losses resulting from which the results of the observations will be ob-
malfunction and damage by construction activities, tained, recorded, digested, and used on the particu-
but also to provide a meaningful picture of the scat- lar project for which the design is being prepared.
ter in results inherent in geotechnics as a conse- Small wonder that the need exists for a book dealing
quence of variations in geology and in construction comprehensively with this subject.
procedures. For example, in the early days of con- RALPH B. PECK-
struction of the Chicago subways, when the loads in
PREFACE
This is intended to be a practical book for use by literature abounds with descriptions of prototype
practitioners. There is information for all those who gadgets that have found little real use in practice. In
plan or implement geotechnical instrumentation selecting information to be included, I have been
programs: owners, project managers, geotechnical guided by the title of the book, and thus there is
engineers, geologists, instrument manufacturers, nothing on geotechnical instrumentation for in situ
specialty geotechnical contractors, civil engineers, measurement of soil and rock properties. Detailed
and technicians. The book should also be helpful to case histories have been excluded in favor of guide-
students and faculty members during graduate lines directed toward the problem-oriented reader.
courses in geotechnical engineering. However, summaries of selected case histories are
A practical book about geotechnical instrumenta- included in Part 5.
tion must go beyond a mere summary of the techni- Finally, a few words about the organization of
cal literature and manufacturers' brochures: it must the book and how it may be used. The book is di-
hold the hands of readers and guide them along the vided into seven parts, each with a self-explanatory
way. This need has created two difficulties for me. title. Readers looking for an overview may start
First, my own practical experience is that of one with the Foreword and Chapter 1, then scan
person and does not arm me to write a comprehen- through Chapter 26, The Key to Success. In my
sive guide on my own. I have tried to fill this gap by view, the greatest shortcoming in the state-of-the-
drawing on the experience and opinions of many practice is inadequate planning of monitoring pro-
colleagues, who are identified elsewhere. grams, and therefore problem-oriented readers
Second, it is certain that, soon after publication should give their first concentrated attention to
of this book, I will alter some of my opinions as my Chapter 4, Systematic Approach to Planning Moni-
experience increases. I am well aware that the sub- toring Programs Using Geotechnical Instrumenta-
ject of geotechnical instrumentation is a contentious tion. The various steps in this chapter lead readers
subject, made so by strongly held views among to each of the chapters in Parts 2, 3, and 4: Chapter
practitioners and by vested commercial interests. 4 is therefore the hub of the book. The chapters in
The guidelines in this book are an attempt to convey Part 5, Examples of Instrumentation Applications,
the "best ways" as I see them today. You, the are intended as supplementary chapters to open the
reader, will have your own experience and your minds of readers to the possible role of geotechnical
own best ways, which may differ from mine. I instrumentation 'on various types of construction
therefore have a plea: when you see possibilities for projects and to guide them toward implementation.
improving the content of this book, send me rea- They are not intended as exhaustive summaries,
sonable evidence. My address is in the Directory of state-of-the-art papers, or "cookbooks." If a reader
the American Society of Civil Engineers. Not only uses this book by (1) turning to the chapter in Part 5
will I learn from you, but I will try to disseminate that discusses his or her type of project, (2) noting
the improvements, perhaps ultimately in a second the types of instruments suggested in Part 5, (3)
edition of this book. noting the sketched layouts in Part 5, (4) studying
Length restrictions have strongly influenced the Part 3, Monitoring Methods, for details of the in-
contents. There is no attempt to describe every in- struments, and (5) proceeding with a monitoring
strument, either currently available or described in program, that reader is misusing the book. Turn
published papers. Some available instruments are back to Chapter 4!
not well suited to their intended purpose, and the JOHN DUNNICLIFF
xi
GEOTECHNICAL INSTRUMENTATION
FOR MONITORING
FIELD PERFORMANCE
Every instrument on a project should be
selected and placed to assist with answering
a specific question: if there is no question,
there should be no instrumentation.
Part 1
Introduction
Part 1 is intended to serve as a general introduction. Chapter 1 sets the stage for the
book, describing the role of geotechnical instrumentation and giving a historical
perspective and a look into the future. It is hoped that Chapter 1 will motivate the
reader toward a deeper study of the subject. Chapter 2 presents an overview of key
aspects of soil and rock behavior, targeted for the practitioners who become in-
volved with geotechnical instrumentation programs and who do not have formal
training in soil or rock mechanics.
CHAPTER 1
GEOTECHNICAL INSTRUMENTATION:
AN OVERVIEW
The engineering practice of geotechnical instrumen- The use of geotechnical instrumentation is not
tation involves a marriage between the capabilities merely the selection of instruments but a com-
of measuring instruments and the capabilities of prehensive step-by-step engineering process begin-
people. ning with a definition of the objective and ending
There are two general categories of measuring with implementation of the data. Each step is criti-
instruments. The first category is used for in situ cal to the success or failure of the entire program,
determination of soil or rock properties, for ex- and the engineering process involves combining the
ample, strength, compressibility, and permeability, capabilities of instruments and people.
normally during the design phase of a project. Ex-
amples are shown in Figure 1.1. The second cate-
gory is used for monitoring performance, normally 1.2. WHY DO WE NEED TO MONITOR
during the construction or operation phase of a proj- FIELD PERFORMANCE?
ect, and may involve measurement of groundwater
pressure, total stress, deformation, load, or strain. The term geotechnical construction can be used for
Examples are shown in Figure 1.2. This book is construction requiring consideration of the en-
concerned only with the second category. gineering properties of soil or rock. In the design of
During the past few decades, manufacturers of a surface facility, the ability of the ground to sup-
geotechnical instrumentation have developed a port the construction must be considered. In the
large assortment of valuable and versatile products design of a subsurface facility, consideration must
for the monitoring of geotechnically related parame- also be given to the ability of the ground to support
ters. Those unfamiliar with instrumentation might itself or be supported by other means. In both
believe that obtaining needed information entails cases, the engineering properties of the soil or
nothing more than pulling an instrument from a rock are the factors of interest. The designer of
shelf, installing it, and taking readings. Although geotechnical construction works with a wide vari-
successful utilization may at first appear simple and ety of naturally occurring heterogeneous materials,
straightforward, considerable engineering and plan- which may be altered to make them more suitable,
ning are required to obtain the desired end results. but exact numerical values of their engineering
3
,<)
,b, ,d,
figure 1.1. wmpks of measuring instrumenU IOf in situ determin<ltion of soil or rock properties: (ol)
Pie:zocone: combined static cone ,lind pore pressure probe (courtesy of Geotechniques International,
Inc., Middleton, MAl; (b) V.lne sM;1Ir equipment (courtesy of Ceonor AIS, Oslo, Norway); Ie) self-boring
pressu~ter (courtesy of Cambridg~ Insilu, CamtN-idse. Engl.lnd);.OO (d) borehole deformation gage
(courtesy of Geokon, Inc., lebanon, NH).
4
WHERE ARE WE NOW? 5
properties cannot be assigned. Laboratory or field instrument procurement on a low-bid basis will re-
tests may be performed on selected samples to ob- main a stumbling block to good field performance.
tain values for engineering properties, but these
tests will only provide a range of possible values.
The significance of these statements about 1.5. WHERE HAVE WE BEEN?
geotechnical construction can be demonstrated by
comparison with steel construction. A designer of a Figures 1.3-1.15 show examples of past uses of
steel structure works with manufactured materials. geotechnical instrumentation.
The materials are specified, their manufacture is The birth of geotechnical instrumentation, as a
controlled, and fairly exact numerical values of en- tool to assist with field observations, occurred in
gineering properties are available for design. An ac- the 1930s and 1940s. During the first 50 years of its
curate analysis can be made and design plans and life, a general trend can be observed. In the early
specifications prepared. Then, provided construc- years, simple mechanical and hydraulic instruments
tion is in accordance with those plans, the structure predominated, and most instrumentation programs
will perform as designed. There will generally be no were in the hands of diligent engineers who had a
need to monitor field performance. Similar remarks clear sense of purpose and the motivation to make
apply to reinforced concrete. In contrast, the design the programs succeed. There were successes and
of geotechnical construction will be based on judg- failures, but the marriage between instruments and
ment in selecting the most probable values within people was generally sound. In more recent years,
the ranges of possible values for engineering prop- as technology has advanced and the role of geo-
erties. As construction progresses and geotechnical technical instrumentation has become more secure,
conditions are observed or behavior monitored, the more complex devices with electrical and pneu-
design judgments can be evaluated and, if neces- matic transducers have become commonplace.
sary, updated. Thus, engineering observations dur- Some of these devices have performed well, while
ing geotechnical construction are often an integral others have not. At the same time, the technology
part of the design process, and geotechnical in- has attracted an increasingly large proportion of the
strumentation is a tool to assist with these obser- geotechnical engineering profession, and an in-
vations. creasing number of instrumentation programs have
been in the hands of people with incomplete motiva-
tion and sense of purpose. There have continued to
1.3. WHAT CAPABILITIES MUST THE be successes and failures but, in contrast to the
PEOPLE HAVE? early years, a significant number of the failures can
be attributed to an unsound marriage between in-
Basic capabilities required for instrumentation per- struments and people.
sonnel are reliability and patience, perseverance, a
background in the fundamentals of geotechnical en-
gineering, mechanical and electrical ability, atten- 1.6. WHERE ARE WE NOW?
tion to detail, and a high degree of motivation.
The state of the art of instrument design is now far
ahead of the state of the practice by users, and
1.4. WHAT CAPABILITIES MUST THE many more imperfections in current instrumenta-
INSTRUMENTS HAVE? tion programs result from user-caused people prob-
lems rather than from manufacturer-caused instru-
Reliability is the overriding desirable capability for ment problems. As users we are fortunate in having
instruments. Inherent in reliability is maximum access to such a wide variety of good instruments.
simplicity, and in general the order of decreasing It is our responsibility to develop an adequate level
simplicity and reliability is optical, mechanical, of understanding of the instruments that we select
hydraulic, pneumatic, electrical. Also inherent in and to maximize the quality of our own work if we
reliability is maximum quality. Lowest cost of an are to take full advantage of instrumentation tech-
instrument is rarely a valid reason for its choice, nology. The greatest shortcoming in the state of the
and unless high quality can be specified adequately, practice is failure to plan monitoring programs in a
(bl
(.,
«,
(d)
(.,
" ..
(f)
Fillure 1.2. Enmpla of mnSllrins instruments for m<woitorinll flCld perlorTJUlRCt: (ill) twin-luM hr-
dr"l.Ilic piezometer (coortesy of CMIKhnkaJ Instruments (U.K.) ltd., leamington SpJi, Engl,lnd); (bl
vibr.ting win! pie:zomete... (cour1~ of TtlemK, Asn~rti, Fr.nct); Ie) vibrililinl wire slressmeler
(courtesy of Geokon, Inc., ~non. NH); (d) ~ cell (courtesy of Proceq 5A, Zurich, 5wilzm.nd);
It) embedment t.llrth prnstJre celt (courtesy of Thof Intenwotioml, loc., Seatde, WA); (0 surface-
mounted vibtatins wire strain PIe (cou rtesy of 'rad Gage, a Division of Klein As.soci.JIles, Inc., Salem,
NH); (g) multipoint fiKed borehole exlensometer (courtesy of Soillnslrumenls ltd., Ucklield, England);
;lIId (h) inclinometer (courtesy of Slope Indicator Company, Se.JIule, WA).
•
,h)
7
8 GEOTKHNICAL INSTRUMENTATION: AN OVERVIEW
figure 1.4 . Delennination of load in a steel strut, using a mKh.ani- Figur~ 1.6. Ilt§tollling fixed em~nkment extensometer with vibr.t-
cal str.llin gage. Open cui for sta tion in day. Chiago Subw.JIy, ing wire tr.nsducer. Balderkead Dam, England, 1963 (courtesy of
1948 (courtesy of R.J.1!)h B. Peck). Arthur D. M. Penm.Jn).
WHERE ARE WE NOW~ ,
Figure 1.7. Manomeler panels for twin-tube hydraulic piezometers. Plover Cove Main Dam, Hong
ICong, 1965 (.lifter Dunnicliff. 1968). Reprinted by permission of tnstitulion of Civil Engioeers, London.
f,&ure I .B. Inst.lllling fixed embankment u tensometers in em- figure 1.9. Bonded resistance strain &-lgts on ~ted steel
boIInkment cbm. ludinglon Pumped Storage Projed, ludington, liner for sofl grtltmd tunnel. POr1 Richmond Woller Pollution Con-
MI,1 972. lrol Project, Slalen Island, NY, 1974.
10 GEOTECHNICAL INSTRUMENTA.TlON: AN OVERVIEW
Many practitioners who become involved with ments of rocks or minerals that have not been al-
geotechnical instrumentation programs do not have tered by chemical decomposition. Inorganic silt is a
formal training in soil or rock mechanics. The pur- fine-grained soil with little or no plasticity and can
pose of this chapter is to present a brief and simple generally be classified as cohesionless. Organic silt
overview of the key aspects of soil and rock behav- is a fine-grained soil with an admixture of organic
ior that relate to the use of geotechnical instrumen- particles and behaves as a plastic cohesive soil.
tation. For a thorough treatment of soil behavior Clay is a cohesive soil consisting of microscopic
readers are referred to Holtz and Kovacs (1981) and and submicroscopic particles derived from the
Terzaghi and Peck (1967). McCarthy (1977) pres- chemical decomposition of rock constituents.
ents similar material oriented for students at techni-
cal colleges. Rock behavior is well described by 2.1.3. Stress and Pressure
Blyth and DeFreitas (1974) and Franklin (1988).
Stress and pressure are defined as force per unit
area, with typical units of pounds per square inch
2.1. BEHAVIOR OF SOil Ob/in.2) or pascals (Pa). Strictly, pressure is a gen-
eral term meaning force per unit area, and stress is
2.1.1. Constituents of Soil the force per unit area that exists within a mass.
Soil is composed of solid particles with intervening However, in geotechnical engineering, the terms
spaces. As shown in Figure 2.1, the particles are
referred to as the mineral skeleton and the spaces as
pore spaces, pores, or voids. The pore spaces are
usually filled with air andlor water. A soil in which
the pore spaces are completely filled with water is
called a saturated soil. If any gas is present in the ~~~"..::~lH!b~~4--.., Mineral
pore spaces, the soil is called an unsaturated soil. skeleton
13
14 BEHAVIOR OF SOIL AND ROCK
Fine
Wtlltr orifice. 10lb
surface Porous with
piston Pinon valve
.'.:. . Cont.iner
. Sa turated
soil..
L_ _ -""=--_-'=', Spring
with
Iti llness
101blin.
Idl 1.1
figure 2.3. Spring ;analogy for soil behavior.
'"
BEHAVIOR OF SOIL 15
deflection of 1 inch. In Figure 2.3c, a 10-pound force total stress = effective stress + pore water pressure.
has been applied to the piston. The water is not free
to escape; therefore, the spring cannot compress This is Terzaghi's principle of effective stress. The
and cannot carry the newly applied force. The wa- following symbols are normally used:
ter must therefore carryall the force, and the pres-
sure gage will show an increase immediately as the Total stress, (J'
force is applied. If the valve is now opened, water Effective stress, (J"
will pass through the orifice and the piston will de- Pore water pressure, u
scend. Figures 2.3d and 2.3e show intermediate
steps, and Figure 2.3f shows the condition when the Thus,
piston has descended 1 inch and there is no further
flow of water. Because the spring has now been (J' = (J" + u.
compressed 1 inch, it must be carrying a force of 10
pounds. The spring is now carrying all the force, Forces and stresses are plotted in Figure 2.4a. It
and the pressure gage has returned to the same can be seen from the figure that the rates of pressure
reading as in Figure 2.3b. Table 2.1 summarizes the change decrease as time increases: this is consistent
steps and shows the sharing of applied force be- with the observation that the flow of water through
tween the spring and water. It can be seen from the the orifice in the piston decreases as the water pres-
table that the sum of the forces carried by the spring sure in the container decreases.
and the water is always equal to the force on the
piston.
2.1.6. Consolidation
Effective stress is defined as the force acting be-
tween the points of the mineral skeleton per total The process of gradual squeezing out of water, with
area. Because a cross-sectional area of 1 square the accompanying transfer of total stress to effec-
inch has been chosen in the above analogy, all the tive stress and decrease in pore water pressure, is
forces in Table 2.1 are numerically equal to stresses called consolidation. Figure 2.4b shows the vol-
in Ib/in. 2 if a real soil is considered. By thinking now ume change that occurs during consolidation. The
in terms of stresses, it can be seen that the force on amount by which the pore water pressure exceeds
the piston represents the total stress, the force car- the equilibrium pore water pressure is called the
ried by the spring represents the effective stress, excess pore water pressure, and the gradual de-
and the force carried by the water represents the crease of this pressure is often referred to as dissi-
pore water pressure. The following relationship al- pation of pore water pressure.
ways applies: As a practical example of consolidation, consider
,. BEHAVIOR OF SOIL AND ROCK
Force carried by waler, or pore stress in a particular soil , the greater is its shear
witer pressure
strength. The shear st rength is a measure of the
Force carried by spring. Or
effective $tre5S in soil resistance to sliding between grains that are trying
~ 10~~1"--~~:::::o==="------------~ 10 N, to move laterally past each other.
It can now be seen that the gai n in shear strength
~
1o~------~~~------~ ,• during the consolidation process can be monitored
by measuring pore water pressure.
Time
0 "
1.1 2.1.8. Normally Consolidated and
Overconsolidated Soil
A normally consolidated soil is one that has never
been subjected to an effective stress greater than
the existing overburden pressure. Examples include
ocean and lake-bed clays. An overconsolidated soil
is one that has been subjected to an effective stress
greater than the existing overburden pressure. Ex-
amples incl ude clays such as London clay, where
thousands of feet of overbu rden have been eroded.
Timtl
Ibl
2.1.9. Difference Between Pore Water Pressure
figure 2.4. (al Shari ng of applied force and stress. (b) Volume
change. and Groundwater Level
The groundwater level is defined as the upper sur-
face of a body of groundwater at wh ich the pressure
is atmospheric.
a layer of fill for a highway embankment, placed on Figure 2.5 shows three pen orated pipes installed
a clayey founda tion soil. As the fill is placed, pore in a soil within which there is no How of groundwa-
water pressure in the fou ndation immediately in- ter; therefore, groundwater pressure increases uni·
creases and then start s to dissipate, resulti ng in ,~' el fo rmly with depth. When such equilibrium condi-
llement. The rate of settlement depends primarily tions ex ist, the level of water wit hin the pipe will
on the permeability of the fou ndation soil. Perme- ri se to the groundwater level , independent of the
ability is a measure of the rate al which water can location of the pen orations.
move through the soil. Cohesive soils have lower Now consider what happens when a layer of fi ll
permeability than cohcsionlcss soils. and therefore is placed above the sand shown in Figure 2.5 . Fig-
consolidat ion and settlement of cohesive soil s occur
more slowly.
dent that the ability of a soil to resist sliding is re- Figure 2.5. Groundwater level when there is no now of ground-
lated to the effective stress: the larger the effective water.
BEHAVIOR OF SOil 17
Meniscus
-+-----Pa
Point B
(pressure = Pw )
~~~~)!"''''
Point A .:'
force (pressure = Pa) ::.
Shear_-....
(a)
(a)
Shear
force _--...!~~~~~~~~~~
po
(b)
(b)
Figure 2.7. (a) Positive and (b) negative pore water pressure
caused by application of shear force. Figure 2.8. Pore gas and pore water pressure: (a) straw in con-
tainer of water and (b) element of unsaturated soil.
are filled with water, and additional water is pre- Figure 2.8a. The water level in the straw rises to a
vented from entering, pore water pressure de- level higher than in the container and is "held up"
creases. As a practical example, consider the exca- by surface tension forces between the straw and
vation of a slope in overconsolidated clay. Pore water at the meniscus. The pressure in the air is
water pressure decreases as a result of unloading, atmospheric pressure Pa; therefore, the pressure at
but significant additional decrease can be caused by point A must also be P a: if it were not so, there
the development of lateral shear forces. These would be a flow of water to create equality of pres-
shear forces cause deformation, a temporary de- sures at the same level. The pressure at point A
crease in pore water pressure, and a temporary in- must be greater than that at point B (Pw ), because
crease in strength. the pressure in water increases as depth below the
surface increases. P a is therefore greater than P w ,
and the pressure at the air side of the meniscus is
2.1.11. Pore Gas Pressure
greater than the pressure at the water side. It is well
In an unsaturated soil, both gas and water are pres- known that water in a smaller diameter straw rises
ent in the pore spaces, and the pressure in the gas is to a greater height: the pressure difference across
called the pore gas pressure. As with pore water the meniscus is therefore greater, and the radius of
pressure, pore gas pressure act in all directions with curvature of the meniscus is smaller.
equal intensity. Examples of unsaturated soil in- Now consider a meniscus between air and water
clude compacted fills for embankment dams and or- in a pore space within soil, as shown in Figure 2.8b.
ganic soil deposits in which gas is generated as or- The same rule applies as in the analogy: the pore air
ganic material decomposes. When the gas is air, the pressure U a is greater than the pore water pressure
term pore air pressure may be used. U w • The smaller the pore space, the smaller the ra-
The pore gas pressure is always greater than the dius of curvature of the meniscus, and therefore the
pore water pressure. Consider the analogy of a difference between pore air pressure and pore water
straw placed in a container of water, as shown in pressure is greater.
BEHAVIOR OF SOIL 19
2.1.12. Other Terms Relating to Behavior of Soil caused by rearrangement of the relative positions of
grains as shear deformation occurs.
A perched groundwater level is above and not hy-
Shear strength is governed by the nature, size
draulically connected to the more general ground-
and shape of the soil grains, packing density and
water level. For example, groundwater may be
effective stresses within the soil. Shear failure oc-
trapped above a clay layer at shallow depth,
curs when the stresses increase beyond those that
whereas the general groundwater level is in a lower
can be sustained by the soil and the strength is ex-
layer of sand.
ceeded.
An aquifer is a pervious soil or rock stratum that
contains water. An artesian aquifer is confined be-
tween two relatively impervious layers and capable Footing Foundations
of carrying groundwater under pressure. Figure 2.9 depicts a load applied to a footing foun-
The piezometric elevation, or piezometric level, dation located on a soil mass. In the upper figure,
is the elevation to which water will rise in a pipe the load is substantially less than that required to
sealed within the soil, as shown for pipes (c) and (d)
cause failure. The footing transfers the stress to the
in Figure 2.6.
soil, causing the soil to compress in the vertical
direction, and the footing settles. In the lower
figure, a larger load is applied to the footing, stress-
2.1.13. Primary Mechanisms that Control
ing the soil and causing a rotation that tends to push
Behavior of Soil *
the adjacent block of soil out of the ground, as if it
The primary mechanisms that control the engineer- were a rigid body. This movement is opposed by the
ing behavior of soil may be categorized as hy- shearing resistance of the soil along the potential
draulic, stress-deformation, and strength mecha- failure surface. In a cohesionless soil, this shearing
nisms. resistance is frictional, and the magnitude of friction
When a soil is subjected to excess pore water depends directly on the normal stresses against the
pressure, the water flows through the pore spaces in potential failure surface. In a cohesive soil, this
the soil. These may be fairly large, as are the spaces shearing resistance is often purely cohesive, inde-
in openwork gravel, or microscopic, as in the spaces pendent of the normal stresses during the shearing
between the finest clay particles. As it flows, water process but dependent on the history of previous
reacts against the particles, causing friction and a effective stresses.
resistance to flow. The amount of friction depends
on both the velocity of flow and the size of the soil
Deep Foundations
particles. In a very fine-grained soil, there is a larger
area of contact between the water and soil particles Figure 2.10 shows a loaded pile or drilled shaft,
than in a coarse-grained soil, so that there is more embedded in soil. The pile moves downward in re-
friction. The permeability of the soil is governed by sponse to the load and, as it moves with respect to
the amount of friction and resistance to flow. Just as the soil, it mobilizes the skin friction or shearing
the soil acts on the water to retard the flow, so also stress, which resists the sliding along the soil/pile
does the water act on the soil. Water exerts a trac- interface. This accounts for part of the support that
tive force on the soil, in the direction of flow, owing the pile receives from the soil. The rest of the sup-
to the friction. If the soil is saturated, water acts port is from point bearing, the same mechanism that
also with a buoyant force on the soil, whether or not supports a footing on soil.
the water is moving. When soil settles with respect to a pile or drilled
Stress-deformation characteristics of cohesive shaft, the pile is loaded by frictional forces at the
soil are governed by the time required for water interface, causing an increase in stress within the
to flow through the pore spaces in the soil and pile and/or increased settlement. The loading is
for consequent volume changes to occur. For referred to as downdrag or negative skin fric-
cohesionless soils, most of the volume change is tion. Downdrag loading may be caused by several
events. First, when piles are driven through fill
that overlies clay, the loading from the fill may
* Written with the assistance of Norbert O. Schmidt, Professor of cause continuing primary consolidation or second-
Civil Engineering, University of Missouri-Rolla, Rolla, MO. ary time effects. Second, settlements are caused
20 BEHAVIOR OF SOil AND ROCK
So i I co m presses
(a)
~ lower block
(b)
Figure 2.9. Behavior of a footing foundation: (a) settlement of footing and (b) bearing failure.
when ground smface loading is increased after piles Excavated and Natural Slopes
are driven, for example, by placing an approach fill
Excavated and natural slopes often involve layered
alongside a pile-supported bridge abutment. Third,
sediments, which may be parallel to the original sur-
dewatering, site grading, and/or vibration during
face of the slope. One of these layers is usually
pile driving may cause settlement and downdrag
weaker than the others, so the potential surface of
loading.
sliding is noncircular and tends to follow the weak-
,
Load
est layer as depicted in Figure 2.11. When the soil
is relatively homogeneous, the potential surface
of sliding may be more circular. The stability of
slopes in soil is controlled by the ratio between
available shearing resistance along a potential sur-
face of sliding and the shear stress on that surface.
Any increase in pore water pressure along the po-
tential surface of sliding decreases the shearing re-
J
:i3 Shearing resistance sistance and the factor of safety against sliding.
t
n ~
~
~,
opposes sliding
of pile against soil
Retaining Walls
Soil is only partially capable of supporting itself.
Sand requires a retaining structure if it is to stand
vertically, but clay may support itself vertically for
Development of limited heights for some period of time. Figure 2.12
"point" bearing shows a retaining wall with forces against it. The
Figure 2.10. Behavior of a driven pile or drilled shaft. earth pressure shown pushing the wall towards the
BEHAVIOR OF SOil 21
o·
Force caused
. . . . - by active earth
~ pressure pushing
.' on wall
o·
Force caused
by passive ... :c:.:. p'. " ()' '.0. o.
earth pressure _ _
•I ----- Shearing resistance
f f f f Bearing stress
opposes sliding
resistance
opposes holding wall up
movement
Figure 2.12. Behavior of a retaining wall.
Figure 2.11. Failure of a slope in layered soil.
left is that portion of the pressure that remains after tion, shown in Figure 2.9, the soil behaves as if it
the soil tries to support itself. It is known as the were a rigid body. Rotation is resisted by mobiliza-
active earth pressure and is reduced to its minimum tion of shearing resistance along the arc, as de-
value when the wall slides and tilts slightly to the scribed for footing foundations.
left. The sliding also mobilizes the shearing resis-
tance between the soil and the base of the wall. In Embankments on Soft Ground
front of the wall on the left, the passive earth pres-
The behavior of embankments on soft ground tends
sure is mobilized as the soil resists sliding.
to be dominated by the properties of the soft
ground. A potential circular failure surface may de-
Braced Excavations
velop, with a large portion of the surface in the
When compared with retaining walls, braced exca- weak foundation material as shown in Figure 2.14a.
vations differ in their mobilization of earth resis- However, the loading of the embankment may
tance because significant sliding or tilting toward cause settlement and lateral bulging of the founda-
the excavation is not permitted. The wall of a tion, as shown in Figure 2.14b, long before the rota-
braced excavation may be braced against the oppo- tional failure occurs. The lateral bulging of the soft
site wall with cross-lot bracing, braced against the ground transfers horizontal tension to the embank-
bottom or the adjacent wall with rakers, or held in ment, which may experience tension cracking,
place by tieback anchors that pull against the soil since it is less deformable than the soft foundation.
outside the excavation. Stresses against a braced
wall are greater than against a retaining wall be- Embankment Dams
cause the passive pressure is greater. However, if
Embankment dams may experience distress or fail-
bracing is properly installed, deformation of the soil
ure in a variety of ways.
is small and buildings on adjacent sites may there-
Overtopping may result from incorrect estima-
fore experience only small settlements and little dis-
tress. When a deep braced excavation is made in
soft soil, soil tends to move into the bottom of the
excavation: this is called bottom heave, and special
construction techniques may be required to control
deformation.
Embankments
Shearing
resistance
Embankments of relatively homogeneous soils opposes
overlying hard ground tend to fail by rotation along movement
an almost circular arc. This is depicted in Figure Figure 2.13. Behavior of an embankment of relatively homoge-
2.13. As with the bearing failure of a footing found a- neous soil.
22 BEHAVIOR OF SOil AND ROCK
Original elevation
of crest
After
Transverse
crack
(a)
~ ~
¢:J ¢:l {J, Q c:::>
Deformation
cially silty sand, an earthquake can cause the foun-
(b)
dation to liquify and flow, as with the near failure of
Figure 2.14. Behavior of an embankment on soft ground: (a) rota- the San Fernando Dam in California.
tional slide along arc and (b) settlement and lateral bulging of soft Even if the design of the dam is adequate, the
foundation. weight of the embankment dam on the underlying
soil or rock must be considered. Heavily loaded soil
under the dam may settle, and there will be down-
ward and lateral movements of the base of the
tion of storm water volume or duration. It may also
dam. Moreover, even well-compacted fill material
be the result of slope failure on the reservoir rim,
will experience settlements when loaded with over-
causing a large volume of material to slide into the
lying material. and poor compaction procedures
reservoir and induce a wave, as occurred at Viaont
will result in greater settlements. If the crest of the
Dam in Italy.
dam is initially level, with time it will settle, and the
Internal erosion, or piping, can cause distress to
center of the dam will settle the most. If the abut-
an embankment dam. Erosion occurs where seep-
ments are steep, the settlements may put the crest
age water is flowing at a velocity sufficient to carry
of the dam in tension, as shown in Figure 2.15, pos-
soil particles along with the water. If the soil has
sibly causing cracks transverse to the axis of the
some measure of cohesion, a pipe or small tunnel
dam.
may form at the downstream exit of the seepage
path as the soil is eroded. As the pipe lengthens in-
Soft Ground Tunnels
side the embankment, there is less resistance to
flow, the flow in the pipe increases, and piping ac- When a tunnel is excavated in soft ground, it may
celerates. Silts and fine sands are most prone to be excavated to a diameter slightly larger than its
piping, because they erode easily, and when moist lining so that the lining may be placed, and the
are sufficiently cohesive to form the walls of a pipe soil may be unsupported for a short time period.
without collapsing. In a well-designed embankment Stresses are relieved and soil tends to move inwards
dam, piping is prevented by sizing downstream ma- toward the tunnel cross section as well as into the
terial such that the pore spaces of that material are face. Although the space between the soil and lining
just smaller than the larger sizes of the mineral is usually small and may be grouted, the ground
skeleton in the adjacent upstream material, so that surface may settle, with greatest settlements occur-
migration of particles is blocked. Such a down- ring directly over the tunnel as shown in Figure
stream material is known as a./ilter. 2.16.
When the slopes of a dam are too steep, the dam If the groundwater level was previously above
may fail as described above in the discussion on the invert of the tunnel, the tunnel may drain the
embankments. Many river sediments consist of soft soil. In granular soils, the groundwater may have to
ground, and a dam constructed over these materials be predrained to the level, of the invert, so that the
may behave as shown in Figure 2.14. tunnel can be constructed. The reduction in ground-
If a dam is built on loose granular material, espe- water level will increase effective stresses in the soil
BEHAVIOR OF ROCK 23
deposit and form planes of weakness. Flow banding geologist yet are critical to the mechanical character
is a characteristic of igneous volcanic rocks such as and behavior of the rock mass.
basalt, in which layers of lava are often separated On the small scale, the engineer has to consider
by volcanic froth (pumice stone) and windborne or how mineral composition and microtexture affect
water-deposited layers of ash or bentonite clay. properties such as rock strength and durability.
Tension joints (cracks) have been formed by con- Pores are by far the weakest rock-forming compo-
traction of igneous rocks during cooling and of nent, and porous rocks are much weaker and more
sedimentary rocks during the accumulation of over- deformable than dense ones. Porosity (the ratio of
lying material. pore volume to total volume) can exceed 50% in
Secondary structures have been generated by the some sandstones and highly weathered granites.
subsequent action of metamorphism, heat, and Hard rocks (competent rocks) are usually formed
pressure, caused mainly by drifting of rigid conti- from hard and resistant minerals. Soft rocks such as
nental blocks on the more fluid mantle and core of shales are usually composed of clay and similar
the earth. Slowly but continuously the continents minerals that are themselves weak and deformable.
are being thrust one beneath the other, in a process These rocks often break down when subjected to
of continental drift that is creating mountain chains wetting and drying because their minerals tend to
and belts of volcanic and earthquake activity along attract a boundary layer of water. Limestones com-
the lines of impact. This has resulted in the folding posed of calcite and salt rocks composed of saline
and sometimes complete overturning of beds that minerals, although quite strong and brittle when
were originally horizontal. The same forces acting found in near-surface engineering construction, de-
on more brittle rocks have created not only further form and flow quite readily when subjected to the
tension joints but also faults, defined as surfaces of higher temperature and pressures found in deep
sliding and shearing. Fault zones are common, con- underground mines.
sisting of several subparallel faults or shears (small Mineral composition has an important influence
faults) with interposed fault breccia (broken rock) on how rocks deform in response to stress. Those
and fault gouge (finely crushed rock and clay). like granite that are composed mainly of hard min-
erals are usually elastic, because when moderate
levels of stress are applied and then removed the
2.2.2. Engineers' View of Rock
rocks return to their original size and shape. If the
The many hundreds of rock names assigned by small deformations that occur are in direct pro-
geologists often mean little in the context of an en- portion to the magnitude of the applied stresses,
gineering project. The engineer is concerned not so the rock is described as linearly elastic. At higher
much with the history of the rock or its precise stresses and after very little deformation, they fail
mineral composition as with its potential behavior suddenly in a brittle manner. Rocks like potash that
in an excavation or foundation. When classifying contain soft minerals are often inelastic and may
rocks for engineering purposes, the engineer must creep (continue to distort) when a high level of
consider properties that often go unreported by the stress is maintained. This is termed viscous or vis-
BEHAVIOR OF ROCK 25
E
u
OJ
N
'v;
""-
u
.2
CIl
Typical
rock block
Discontinu ities,
Rock mass
set 2
( macrostructure)
coelastic behavior. Their failure is ductile (gradual Figure 2.19. Size-strength rock mass classification diagram (after
and accompanied by large deformations). Franklin, 1986).
Rocks with subparallel platy minerals, as noted
earlier, tend to be fissile and are also anisotropic
(with strengths that vary according to the direction
of loading). Slates are characterized by slaty cleav- and the terms jointing and jointed (e.g., widely or
age so that they can be split readily into thin plates, closely jointed) are often used to apply to discon-
and schists and gneisses by foliation surfaces along tinuities irrespective of their origin. Nearly all as-
which platy mica minerals are abundant. In general, pects of rock behavior, both at the surface and
such rocks are much weaker than those with ran- underground, are controlled much more by the
domly oriented, equidimensional, and interlocking characteristics of such discontinuities than by those
grains. of the intact rock blocks within the mass.
Even more important in an engineering context Figure 2.19 shows a multipurpose size-strength
than minerals and microtexture are the features of rock mass classification that takes into account the
rock on a larger scale. Rock engineers distinguish two features of the rock mass that are usually the
between intact rock and rock mass. The properties most important in engineering applications: the size
of intact rock, otherwise termed rock material, are of rock blocks and their strength, plotted respec-
those that can be measured by testing a small speci- tively along the vertical and horizontal axes. Zones
men of solid rock in the laboratory. The same prop- in the rock mass that plot toward the lower left of
erties measured on the scale of the rock mass are the diagram are broken and weak and easy to exca-
affected by large-scale structural features and can vate but difficult to support in excavations. Those
only be measured directly by large-scale in situ toward the upper right are massive and strong and
tests. Figure 2.18 shows the difference between in- require blasting for excavation but are often self-
tact rock and rock mass. The r:ock mass is nearly supporting. The diagram is contoured to give a rock
always much weaker, more deformable, and more quality index, which is a combined measure of
permeable than the intact material it contains. This block size and strength.
is because of the presence of discontinuities, a Note must be taken of other discontinuity char-
global term used for the various types of fracture acteristics that affect rock mass character and be-
and weakness planes discussed earlier: joints, havior, such as the orientation (dip magnitude and
faults, shears, bedding planes, and surfaces of direction) of joint sets. In a rock slope, for example,
cleavage and foliation. Joints are the most common, horizontal jointing generally has little effect on sta-
26 BEHAVIOR OF SOIL AND ROCK
bility, whereas joints that dip steeply into the exca- 2.2.4. Stress in Rock
vation may well be a source of sliding failures.
In contrast to soils, which are usually deformable,
Other characteristics of discontinuities such as
rocks are rigid and permit the transfer and storage
roughness, continuity, and the presence or absence
of high stresses. Stability of a rock mass therefore
of a clay infilling determine the shear strength of the
depends on three principal factors: the characteris-
discontinuity surfaces. Roughjoints with no infilling
tic of the intact rock and particularly of the discon-
are strong compared with smooth, slickensided,
tinuities, the characteristics of the groundwater
and clay-filled fault planes.
regime, and the magnitudes of ground stress in rela-
tion to rock strength and brittleness.
The stresses acting vertically at depth can be es-
2.2.3. Water in Rock
timated, as for soils, from the weight of overlying
Seepage through rock masses occurs almost exclu- materials. The magnitudes of horizontal stresses
sively along discontinuities, with little or no flow cannot be estimated in this way and can be much
through the intact rock blocks unless these are ex- greater than those acting vertically. At some loca-
tremely porous. Discontinuity characteristics gov- tions, even close to the surface, horizontal stresses
ern water pressures and flow rates through the rock can be ten or more times greater than vertical
mass, just as they govern mechanical characteris- stresses. High horizontal stresses close to the
tics such as rock mass strength and deformability. surface can result in buckling and heaving of shal-
Flow and pressures are governed by the aperture low excavations and in squeezing and cracking of
(openness or tightness) of the discontinuities and by tunnel linings and buried concrete pipes. When high
their spacing and continuity. stresses surround an underground excavation, they
Joint water pressures within a rock mass can can be a stabilizing influence, helping to hold in
have a most important influence on stability. When place the rock arch above a tunnel, a mine stope, or
there is a need to monitor these pressures, measure- an underground powerhouse cavern. However, if
ments must be made within zones of open jointing. excessive they can cause extensive damage in the
The principle of effective stress discussed for soil in form of squeezing or rockbursting.
Section 2.1.5 applies equally to the rock mass, ex- Rock engineers distinguish between virgin
cept that the pore water and mineral skeleton sys- stresses (also called in situ stresses) created by geo-
tem are replaced by the joint water and rock block logic processes and induced stresses caused by ex-
system. cavation. We can speculate that the virgin stresses
In strong and durable rock masses, high joint were first established by continental drift as dis-
water pressures are the main problem related to cussed in Section 2.2.1. They have since been
groundwater, and these can usually be relieved ef- modified by the deposition and erosion of geologic
fectively by drainage. In weaker and less durable materials. In the past, many locations on the sur-
rock types such as shales, further problems may face of the earth were overlain by several miles of
result from swelling, and rocks containing clay min- soil, rock, or ice. Geologic weathering and erosion
erals are particularly prone to breakdown. Closely over many millions of years have removed this
jointed or less durable rock types are also suscepti- overburden. However, the horizontal stresses ex-
ble to internal erosion within the jointing or to ex- isting in rocks that were once deeply buried may not
ternal erosion by processes of raveling and slaking. be entirely relieved by erosion. Rigid rock types, in
Vertical rock cliffs are often undercut by erosion, particular, often contain large virgin horizontal
and groundwater can be a hazard also in tunnels and stresses that reflect their geologic history and that
underground chambers. Clayey fault gouge acts as a have important influences on present-day engineer-
barrier to high-pressure water. Tunnels penetrating ing construction.
such faults can encounter sudden and catastrophic When an excavation is made in rock, the stresses
inflows of water and broken rock. originally carried by the excavated material are
In other respects, the groundwater regime in transferred to the rock that remains in place. Stress
rocks is similar to that found in soils, so that terms levels around tunnels are amplified to several times
such as groundwater level, perched groundwater their virgin values. The rocks may become over-
level, aquifer, artesian aquifer, and piezometric ele- stressed and may suffer from squeezing or explo-
vation and level, defined in Section 2.1, apply with sive bursting, depending on whether the materials
the same definitions in a rock engineering context. are comparatively ductile or brittle.
BEHAVIOR OF ROCK 27
Figure 2.20. Primary mechanisms of rock slope failure (after Hoek and Bray, 1974). By permission of
the Institution of Mining and Metallurgy.
Gravitational failure
(caving and raveling)
Rockburst
---r' -+ 4- 4---
Figure 2.21. Primary mechanisms of yield and failure of underground openings in rock.
ing, the methods used for excavation (e.g., blasting intact rock is exceeded by the magnitudes of exca-
or boring machine), and the type and quality of sup- vation-induced stresses. Classic examples occur in
port installed. Each of these mechanisms, shown in South African mines at depths approaching 2-3
Figure 2.21, is characteristically different and calls miles (3-5 km) where even the high-strength
for its own individual approach to monitoring and rocks are incapable of sustaining the stresses that
design of instrumentation systems. develop.
Gravitational failures (caving and raveling Squeezing behavior is characteristic of rock
mechanisms) occur when rock blocks fall from the masses that are overstressed but relatively weak or
roof of an underground opening without appreciable closely jointed. The rock mass yields by shearing
breakage of intact rock material. Such failures are along weak discontinuity surfaces or by crushing or
governed entirely by discontinuities. They start distortion of the rock blocks. Tunnel closures in
with the fall of small keystone blocks and progress excess of 3 ft (1 m) have been recorded, for ex-
upward until they either reach the ground surface or ample, by convergence monitoring in the Austrian
are arrested by bulking of broken rock within the Alps.
opening (broken rock occupies a greater volume Swelling behavior is encountered when the min-
than undisturbed rock). Uncontrolled caving in a erals of the rock are themselves of a swelling vari-
Quebec mine in 1980 migrated upward until it ety. Clay minerals are the most common in this cat-
reached the saturated overburden soils, which egory. Some shales contain montmorillonite and
liquified and flowed into the workings. Miners were similar clay minerals that adsorb water, expand,
killed, and the affected levels of the mine had to be and eventually generate high pressures on tunnel
closed. liners and support systems. Swelling problems are
Rockbursting is the characteristic mode of failure also experienced in anhydrite rock types that con-
in brittle rock materials where the strength of the vert to gypsum when wetted.
BEHAVIOR OF ROCK 29
Stress relief closure occurs most commonly in riods, sometimes of years or decades. Hydroelec-
openings excavated in horizontally bedded sedi- tric turbine pits in Niagara have suffered from stress
mentary rocks. The beds expand as a result of relief closure. The inward movement in one pit has
stress relief and slide inward, exerting very high been monitored monthly since 1905, and total
pressures on liners and supports. Some sliding oc- movement has exceeded 4 in. (100 mm), sufficient
curs immediately when the rock is blasted, but con- to cause severe misalignment problems and shatter-
tinuing creep along the bedding planes results in the ing of a cast iron support strut.
development of squeeze over much longer time pe-
560 REFERENCES
Walton, R. J., and S. M. Matthews (1978), "Measurement of Stress Cell Measurements in Soil," J. Geotech. Eng. Div.
Rock Deformation, Stress Changes and Rock Competency ASCE, Vol. 108, No. GT12, Dec., pp. 1529-1548.
in the 18CC/12CZ2 Crown Pillar at the CSA Mine Cobar- Weir-Jones, I., and T. G. Bumala (1975), "The Design of Slope
Part One," Support and Stabilization of Stopes, Proj. Rep. Stability Monitoring Systems," Soc. of Mining Engineers,
No. 33, CSIRO Division of Applied Geomechanics, Aus- AIME, Fall Meeting, Salt Lake City, UT, Sept., Preprint
tralia. No. 75-AM-338.
Walton, R. J., and G. Worotnicki (1978), "Rock Stress Measure- Weltman, A. J. (1977), "Integrity Testing of Piles: A Review,"
ments in the 18CC/12CZ2 Crown Pillar Area of the CSA Rep. PG4, Construction Industry Research and Information
Mine, Cobar, N.S.W.," Proj. Rep. No. 38, CSIRO Division Association (C.I.R.I.A.), London.
of Applied Geomechanics, Australia. Whitaker, T. (1964), "Load Cells for Measuring the Base Loads
Walton, R. J., and G. Worotnicki (1986), "A Comparison of in Bored Piles and Cylinder Foundations," Current Paper,
Three Borehole Instruments for Monitoring the Change of Eng. Ser. No. 11, Building Research Station, Watford,
Rock Stress with Time," in Proceedings of the International England.
Symposium on Rock Stress and Rock Stress Measurements, Whitaker, T., and R. W. Cooke (1966), "An Investigation of the
Stockholm, Centek Publishers, Lulea, Sweden, pp. 479- Shaft and Base Resistances of Large Bored Piles in London
488. Clay," in Proceedings of the Symposium on Large Bored
Ward, W. H., J. B. Burland, and R. W. Gallois (1968), "Geo- Piles, Institution of Civil Engineers, London, pp. 7-49.
technical Assessment of a Site at Mundford, Norfolk, for a Whittaker, B. N., and G. J. M. Woodrow (1977), "The Constant-
Large Proton Accelerator," Geotechnique, Vol. 18, No.4, Tension Strain Wire Borehole Extensometer and Its Appli-
pp. 399-431. cation to Instrumentation of Underground Openings," in
Ward, W. H., D. J. Coats, and P. Tedd (1976), "Performance of Proceedings of the International Symposium on Field Mea-
Tunnel Support Systems in the Four Fathom Mudstone," surements in Rock Mechanics, Ziirich, K. Kovari (Ed.),
Current Paper 25176, Building Research Establishment, Balkema, Rotterdam, Vol. 1, pp. 437-448.
Watford, England, Mar. Wightman, W. D., S. J. Calabrese, and E. L. Foster (1980),
Ward, W. H., P. Tedd, and N. S. M. Berry (1983), "The Kielder "Special Study of Precast Concrete Tunnel Liner Demon-
Experimental Test Tunnel: Final Results," Geotechnique, stration, Lexington Market Tunnels, Baltimore, Maryland,"
Vol. 33, No.3, pp. 275-291. U.S. Department of Transportation, Urban Mass Transpor-
Wardlaw, E. G., L. A. Cooley, A. E. Templeton, and R. L. tation Administration, Rep. No. UMTA-MD-06-0039-81-1.
Fleming, Jr. (1984), "Slide Stabilization with Gravel Wilen, B. 0., and W. P. MacConnell (1973), "The Modified
Trenches," in Proceedings of the International Conference Fernow Frost Gauge," Soil Sci., Vol. 115, No.4, Apr., pp.
on Case Histories in Geotechnical Engineering, S. Prakash 326-328.
(Ed.), University of Missouri, Rolla, MO, published by Uni- Wilkes, P. F. (1972), "An Induced Failure at a Trial Embank-
versity of Missouri, Rolla, MO, Vol. II, pp. 743-748. ment at King's Lynn, Norfolk, England," in Proceedings of
Warlam, A. A., and E. W. Thomas (1965), "Measurement of the ASCE, Specialty Conference on Performance of Earth
Hydrostatic Uplift Pressure on Spillway Weir with Air and Earth-Supported Structures, Purdue University, Lafay-
Piezometers," in Instruments and Apparatus for Soil and ette, IN, ASCE, New York, Vol. 1, Pt. I, pp. 29-63.
Rock Mechanics, ASTM STP 392, ASTM, Philadelphia, PA, Wilson, A. H. (1%1), "A Laboratory Investigation of a High
pp. 143-151. Modulus Borehole Plug Gage for Measurement of Rock
Warner, J. (1978), "Compaction Grouting-A Significant Case Stress," in Proceedings of the 4th Symposium on Rock Me-
History," J. Geotech. Eng. Div. ASCE, Vol. 104, No. GTI, chanics, Pennsylvania State University, University Park,
July, pp. 837-847. PA, published by the Pennsylvania State University, Uni-
Water Power & Dam Construction (1985), "15th International versity Park, PA, pp. 185-195.
Congress on Large Dams, Part One," Water Power & Dam Wilson, S. D. (1%7), "Investigation of Embankment Perform-
Construction, London, Vol. 37, No.9, Sept., pp. 40-45. ance," J. Soil Mech. & Found. Div. ASCE, Vol. 93, No.
Watson, D. A. (1964), Specifications Writing for Architects and SM4, July, pp. 135-156.
Engineers, McGraw-Hill, New York. Wilson, S. D. (1970), "Observational Data on Ground Move-
Webster, J. L. (1970), "Mica Dam Designed with Special Atten- ments Related to Slope Stability," J. Soil Mech. & Found.
tion to Control of Cracking," in Transactions of the 10th Div. ASCE, Vol. 96, No. SM5, Sept., pp. 1519-1544.
ICOLD, Montreal, International Commission on Large Wilson, S. D. (1973), "Deformation of Earth and Rocklill
Dams, Paris, Vol. 1, pp. 487-510. Dams," in Embankment-Dam Engineering, Casagrande
Weeks, R. C., and P. Starzewski (1986), "Automatic Monitoring Vol., Wiley, New York, pp. 365-417.
and Data Processing," in Proceedings of the U.K. Geotech- Wilson, S. D. (1974), "Landslide Instrumentation for the Min-
nical Conference, Birmingham, England, published by neapolis Freeway," Landslide Instrumentation, Trans. Res.
The Midland Geotechnical Society, Birmingham, England, Record, No. 482, pp. 30-42.
pp. 117-126. Wilson, S. D. (1982), "Instrumentation and Performance of Em-
Weiler, W. A., and F. H. Kulhawy (1978), "Behavior of Stress bankment Dams," Lecture Notes, Nigerian Geotechnical
Cells in Soil," Rep. 78-2, School of Civil and Environmental Association, The Nigerian Society of Engineers, Lagos,
Engineering, Cornell University, Ithaca, NY. Nigeria, Oct. 2.
Weiler, W. A., and F. H. Kulhawy (1982), "Factors Affecting Wilson, S. D. (1984), personal communication, Aug. 29.
REFERENCES 561
Wilson, S. D., and P. E. Mikkelsen (1977), "Foundation In- Wood, L. A., and A. J. Perrin (1984), "Observations of a
strumentation-Inclinometers" (color video tape, film Strutted Diaphragm Wall in London Clay: A Preliminary
and reference manual), U.S. Department of Transportation, Assessment," Geotechnique, Vol. 34, No.4, pp. 563-579.
Federal Highway Administration, Implementation Division, Worotnicki, G., and R. J. Walton (1976), "Triaxial Hollow Inclu-
Office of Research and Development. sion Gauges for Determination of Rock Stresses in Situ," in
Wilson, S. D., and P. E. Mikkelsen (1978), "Field Instrumenta- Proceedings of the Symposium on Investigation of Stresses
tion," in Landslides, Analysis and Control, Transportation in Rock-Advances in Stress Measurement, Institution of
Research Board, Special Rep. 176, Chap. 5. Engineers, Australia, National Conference, Pub. No. 76/4,
Windsor, C. R., and G. Worotnicki (1986), "Monitoring Rein- pp. 1-8.
forced Rock Mass Performance," submitted to International Wroth, C. P. (1975), "In Situ Measurement of Initial Stresses
Symposium on Large Rock Caverns, Helsinki. and Deformation Characteristics," in Proceedings of the
Wissa, A. E. Z., R. T. Martin, and J. E. Garlanger (1975), "The ASCE, Specialty Conference on In Situ Measurement of
Piezometer Probe," in Proceedings of the ASCE, Specialty Soil Properties, North Carolina State University, Raleigh,
Conference on In Situ Measurement of Soil Properties, NC, ASCE, New York, Vol. II, pp. 181-230.
North Carolina State University, Raleigh, NC, ASCE, New Wroth, C. P., and J. M. O. Hughes (1973), "An Instrument for
York, Vol. I, pp. 536-545. the In Situ Measurement of the Properties of Soft Clays," in
Wnuk, S. P., Jr. (1981), "On the Use of Bonded Weldable Strain Proceedings of the 8th International Conference on SM &
Gages for Field Measurements in Hostile Environments," in FE, Moscow, published by USSR National Society for SM
Proceedings of the International Conference on Measure- & FE, Moscow, Vol. 1.2, pp. 487-494.
ments in Hostile Environments, British Society for Strain Yu, T. R. (1983), "Rock Mechanics to Keep a Mine Produc-
Measurements, Edinburgh. tive," Can. Min. J., Vol. 104, No.4, Apr., pp. 61-66.
Wolf, S. (1973), Guide to Electronic Measurements and Labora- Zeigler, E. J., J. L. Wirth, and J. T. Miller (1984), "Slurry
tory Practice, Prentice-Hall, Englewood Cliffs, NJ. Trench Wall Replaces Structure Underpinning," in Pro-
Wolfskill, L. A., and C. Soydemir (1971), "Soil Instrumentation ceedings of the International Conference on Case Histories
for the 1-95 MIT-MDPW Test Embankment," J. Boston in Geotechnical Engineering, S. Prakash (Ed.), University
Soc. Civil Eng., Vol. 58, No.4, pp. 193-229. of Missouri, Rolla, MO, published by University of Mis-
Wolosick, J. R., and A. 1. Feldman (1987), "Reliability of Bend- souri, Rolla, MO, Vol. III, pp. 1287-1296.
ing Moments Backcalculated From Inclinometer Measure-
ments," Geotech. News, Vol. 5, No.3, Sept.