You are on page 1of 352

Block Theory and

Its Application to
Rock Engineering

Prunice.Hall Intenutlorul Sedcs in Civil Ergfunuiry and Engfuuertry Meclu.nics

Williarn J. Hall, Editor


Block Thoor>l and
Its Application to
Rock Enginoering

Richard E. Goodman
Gen-hua Shi
Both of University of California, Berkeley

PRENTICE-HALL, INC., Englewood Cliffs, New Jersey 07632

Library of Congress Cataloging in Publication Data
Goodman, Richard E. (date)
Block theory and its application to rock engineering,
Bibtiography: p.
Includes index.
l. Rock mechanics. 2. Block theory (Rock mechanics)
3. Rock excavation. I. Shi, Gen-hua, 1939-
II. Title.
TA706.c64 1985 624.1',st32 84-3348
rsBN 0-t3-078189-4

Editorial/production supervision and interior design :
Sylvia H. Schmokel
Cover design: Edsal Enterprises
Manufacturing buyer: Anthony Caruso

G)1985 by Richard E. Goodman

All rightsreserved. No part of this book may be
reproduced, in any form or by any means,
without permission in writing from the publisher.

Printed in the United States of America


Prentice-Hall International, Inc., London
Prentice-Hall of Australia Pty. Limited, Sydney
Editora Prentice-Hall do Brasil, Ltda., Ria de faneiro
Prentice-Hall Canada lnc., Toronto
Prentice-Hall of India Private Limited, New Delhi
Prentice-Hall of Japan, Inc., Tokyo
Prentice-Hall of Southeast Asia Pte. Ltd., Singapore
Whitehall Books Limited, Wellington, New Zealand

3u14a4 to rfitstaryll
Suog-as1'3uo1y 'tg
uopuo7' a&ayoS Tnuadutl
,futg apo/1 utlol"re





Excavations 3
Modes of Failure 4
Assumptions of Block TheorY 9
Comparison of Block Theory with Other Analytical
Approaches 11
The Key Block System 19


Equations of Lines and Planes 25
Description of a Block 3l
Angles in Space 40
The Block Pyramid 4l
Equations for Forces 43
Computation of the Sliding Direction 45
ExampleCalculations 48


Contents GRAPHICAL METHODS: STEREOGRAPHIC PROJECTION Types of Projections 57 Stereographic Projection of Lines and Planes 64 Stereographic Projection of a Joint Pyramid 75 Additional Constructions for Stereographic Projection 78 Projection of Sliding Direction 83 Examples 85 Appendix: Important Properties of the Stereographic Projection 92 THE REMOVABILITY OF BTOCKS 98 Types of Blocks 98 Theorem of Finiteness l0l Theorem on the Removability of a Finite.19) to (6.22) 201 . Convex Block 108 Symmetry of Block Types 112 Proofs of Theorems and Further Discussion ll2 Shi's Theorem for the Removability of Nonconvex Blocks l2l JOINT BLOCKS 125 Joint Blocks in Two Dimensions 130 Joint Blocks in Three Dimensions 134 Stereographic Projection Solution for Joint Blocks 138 Computation of Emptiness of Joint Pyramids Using Vectors 147 Applications of Block Theory: An Example 149 6 BLOCK THEORY FOR SUFFICIAL EXCAVATIONS 157 Basic Concepts 157 Conditions for Removability of Blocks Intersecting SurfaceExcavations 166 Identification of Key Blocks Using Stereographic Projection 169 Evaluation of Finiteness and Removabilitv of Blocks Using Vector Methods 179 The Numbers of Blocks of Different Types in a Surface Excavation 183 Procedures for Designing Rock Slopes 188 Removable Blocks in an Excavated Face 198 Appendix: Solution of Simultaneous Equations (6.

Contents tx 7 BLOCK THEORY FOR UNDERGROUND CHAMBERS Key Blocks in the Roof. and Walls 204 Blocks That Are Removable in Two Planes SimultaneouslY: Concave Edges 206 Blocks That Are Removable in Three Planes Simultaneously: Concave Corners 2lI Example: Key Block Analysis for an Underground Chamber 214 choice of Direction for an Underground chamber 225 Intersections of Underground Chambers 232 Pillars between Underground Chambers 236 Comparison of Vector Analysis and Stereographic Projection Methods 238 8 BLOGK THEORY FOR TUNNELS ATTD SHAFTS 2& Geometric Properties of Tunnels 241 Blocks with Curved Faces 244 Tunnel Axis Theorem 249 Types of Blocks in Tunnels 249 The Maximum KeY Block 251 Computation of the Maximum Removable Area Using Vector Analysis 253 Computation of the Maximum Key Block Using Stereographic Projection Methods 271 Removable Blocks of the Portals of Tunnels 279 Appendix: Proofs of Theorems and Derivations of Equations 287 9 THE KINEMATICS AND STABILITY OF REMOVABTE BLOCKS 296 Modes of Sliding 296 The Sliding Force 301 Kinematic Conditions for Lifting and Sliding 304 Vector Solution for the JP Corresponding to a Given Sliding Direction 306 Stereographic Projection for the JP Corresponding to a Given Sliding Direction 307 Comparison of Removability and Mode Analyses 310 Finding the Sliding Direction for a Given JP 310 Mode and Stability Analysis with Varying Direction for the Active Resultant Force 313 Appendix: Proofs of Propositions 325 . Floor.

?t8 xroNl 1G€ sfSNSUSJtU .

^ruuations and in support of foundations are authors*) can be number of examples. C-alifornia 94720. Alternatively. miners. and foundation to show because the threads of the developments are new' it seemed important disbelievers' In the the theoretical foundation for important steps. themselves the end of most chapters. the authors decided to write a book' the practical This work was intended. and in some cases' within the chapters when the material is especially fundamental' The applications of btock theory in planning and design of surface and illustrated in alatge underground . statement. which happened the unfolding of a well-timed inlernational visit. but they are there. Computer programs (available from the the examples used by the reader to duplicate these illustrations. that one could entertain doubts about the universality. the methods are so easily applied. lest there be so simple in their end. it would be the rare reader who could through all in the proper order. and proposi- and rigor of the underpinnings if the complete proofs of theorems of all the proofs tions were not included. Berkeley. to point toward engineers' Yet. applications of interest to tunnelers. *c/o Department of Civil Engineering. it rapidly became obvious that of journal articles' story would require more than the limited space and scope be stated in a Although the succession of development and applications could make his or her way long seiies of articles. Therefore. . and the conclusions completeness.Prefaco by the lur'cky accident As the authors began their collaboration. University of California. right from the beginning. It is not necessaryto follow all the steps in appendices at in order to use the methods of block theory.

The authors wish to thank Ms. a valued colleague suggested that we were then "at the tip of an iceberg. Many colleagues have assisted in the development of this book. The authors have developed a number of computer programs for use with this material. but limitations of space forbad their inclusion here. On first learning of the ideas of block theory. Interested readers are encouraged to write for further information. xtl can be worked using manual stereographic projectiqns. The mathematicat basis of stereo- graphic projection is presented more completely here than in any other book known to the authors. The distinct advantage of computational solution is automatic operation. as presented in some of the references citid. This book is not about stereographic projection or computation as such. William Boyle. certainly speed comprehension. Civil and Environmental Technology pro- gram. since topology and set theory are not standard components of courses in engineering mathematics. Readers should be familiar with matrix notation and vector operations. the level of mathematical skills actually required to understand the material is not advanced. Charles Babendreier. It should be understood that the theory is independent of the methods of its application. Marcia Golner for devoted typing of a laborious manuscript. Dr.) The notation employed here may be somewhat new to many readers. and LapYan Chan. The advantage of stereographic projection is its ease and economy. However. after hearing an introductory lecture. and the book presents all the back- ground required to understand all the developments.. which allows its incorporation in larger enterprises. However. and by the National Science Foundation. We hope that some of you who read this material will discover yet new directions and possibilities in your particular specialties. Goodman Gen-hua Shi . Bernard Amadei. We are especially grateful to Dr. Richard E. Douglas Fuerstenau. but about the geometric facts of intersecting discontinuities penetrating a three-dimensional solid. some familiarity with simple steps in applying -will the stereographic projection. Daniel Salcedo.' We have since exposed considerably more material but have hardly begun to exhaust the possibilities. Partial support was provided by the California Institute for Mining and Mineral Resources Research. (See footnote on pre- vious page. Program Officer. A grant from Horst Eublacher and Associates was also appreciated. Director.

B a plane u(n) The upper half-space of (A)' The matrix transPose of ril. a orBfA closed set AcB .4 is BP Block pyramid contained in B SP Space pyramid (: A+ B I is not a subset of . .B u) The empty set AeB I is an element of set B JP Joint pyramid. EP). the boundary of which (0' 0' 0) passes through the origin xill .{ is parallel to .B . an open set A=B ^B is a subset of A. a closed orB=A set A+B I is not an element of B Excavation pyramid.B A+B B is not a subset of I AIIIB A is parallel and in the -B The complement of .B contains -B AII B . A A+B .{ is a subset of .. the boundary of which (A) passes through the origin (x.4 and normal to 2 and Passes B through Q Av B The union of A and . Q) The upper half-sPace a A unit vector whose boundarY is A^B The intersection of .B the set of elements not arf Dip and dip direction of included in .Notation and Abbreviations A A vector U(ft.B. same direction as .B. y. z) Coordinates of the tiP of L(n) The lower half-space of a vector whose tail is at rC'.4 is not equal to .

Wo pointed into space Ett An exterior edge at the fti Upward unit normal to intersection of two cham- plane i bers AxB Cross product of A and cr* An exterior corner at the B intersection of two cham- bers . W1.xtv Notation and Abbreviations R Radius of the reference A. n The number of joint sets and W. U The unit normal vector Crto The interior corner pointed into a rock block formed by the intersec- w The unit normal vector tion of W1. and.B Scalar product of A and sphere B (dot product) Radius of a circle on the wi Wall i stereographic projection Eti The interior edge formed r Resultant applied force by the intersection of W.

Yet we must cope with hard facts that rhake our task complex. California. In an age oftechnological advance. First. it is doubtful that these architects and builders ever created analytical procedures to govern their engineering activities. rocks differ significantly from most other materials with which we build in possessing numerous flaws and weaknesses that together tend to inter- rupt the continuity of material and divide it into domains of difrerent types. This structure was designed to conduct water along a single bedding plane down'the'limb of a roughly cylindrical fold in quartzite. and creating sculptures in rocks. Unfortunately. crossing joints carried water down to a parallel surface forming the bottom of the folded layer. Common rock displays so many planes of weakness as to be essentially a collec- tion of separate blocks tightly fitted in a three-dimensional mosaic. with precise and powerful analytical tools at our fingertips. 1. it is reasonable to hope for a more fundamental basis for rock engineering. Although ancient civilizations demonstrated remarkable ability in cutting and erecting stone monuments. In this ingenious way the water would flow along a natural surface all the way to the river. a photograph of the spillway for old Don Pedro Dam.1. rock as an engineering material is variable and all-encompassing since we find. Second." An example of how the network of discontinuities in rock affects engi- neering performance of excavations is shown in Fig. We call such material "discontinuous rock. within the earth. The concrete training wall of the spillway can be seen in the upper right-hand corner of the figure. excavating tunnels. chapter I Introductlon Rock is one of the oldest yet one of the least well understood construction materials. erecting fOrtifica- tions. . rock materials possessing all classes of mechanical behav- ior.

and in decomposed . tuffs. however. instead of a smooth water course. Introduction Chap. 1 Flgur€ 1. we assume that the motions of points within individual blocks are derived mainly from rigid body motions of the block system. in which soil-like deformations of the blocks are at least as important as inter- block translations and rotations. hundreds of cubic meters in volume. Iimestonen slate. There are softer rocks. claystones. where the rock material is often considerably stronger than good concrete. consequently. This is true. Further spills are now able to quarry the rock even more effectively. the picture shows a cliff belod the edge of the training wall.1 Rock block eroded from spillway of old Don pedro Dam. and if unchecked. california. quartzite. The system ofcrossjoints and bedding surfaces above and below the uppermost quartzite bed delimited a physical rock block. This is an acceptable assumption in hard rocks such as granite. in Tertiary age sand- stones. gneiss. which was lifted and removed from its position by the running water. This will not happen because the dam has been replaced by a larger structure downstream and the place is now submerged. for example. could destroy the training wall and work backward to release the entire contents of the reservoir. and other rock types. In discontinuous rock. in chalk or very young limestone. But it serves to illustrate how the movement of a kev joint block can create a worsening situation.

disturbed zones ulotrg contacts of different rock types. when a quarry is intended to be converted into a park or development after the product has been removed to a predefined contour. may control the outcome of the work. railroads. shafts and tunnels for subterranean sewer systems. and water power. On the other hand. which may be on the order of a hundred years. open cuts for highways. or from sets in other spatial attitudes. Finally.. This impties that the most immediate applications of this theory will be in rock engineering for underground and surface space. both objectives are combined in one project: for example. by advanced plan- ning. In civil engineering work. important through-going joints. sheared zones. water supply. we say only that the methods described in this book apply to an idealized mode. for example. chambers for hydroelectric turbines and transformers. to shape the interior of an underground gallery. tunnels for roads.r. the structural geologist addresses systems of blocks kilometers on edge formed by systems of major and minor faults and major formational contacts. and mining. As in all systems that have ideal end members. whose validity will always have to be tested in practice. individual ioints parallel to the bedding planes. The objective may be to use the excavated material for fill or in manufac- lor. In any event. open cuts for spillways of reservoirs. Occasionally. in which case the controlling planes of discontinuity arefissures and microfauh. Most of the examples in this book and the main experience of the authors concern excavations in which it is the joint system that decides the rock blocks. The types of discontinuities that chop up the rock mass into blocks depend on the scale of interest. the usual rock mass encountered in excavation work lies somewhere betweenthe extremes considered. EXCAVATIONS An excavation is a new space created by removing earth or soil from its natural place. and other transportation routes. we would not burden you with reading this unless our experiences demonstrated a wide applicability for this theory. and open cuts to site surface structures. the design invari- ably requires that the excavated space remain stable during the process of excavation. the space created byremoval of rock has to behave according to the design. trans- portation routes. and usually for the lifetime of the engineering work. The theory of blocks can be applied to explain fabrics observed in rock specimens. or to gain space at or below the surface. . hydroelectric power.Excavations or altered rocks of any kind. pipelines. water supply. andfaults will bound the individual blocks of the rock mass. In detailed rock excavation. To those who live and work in the real world. Examples of permanent civil engineering excavations are underground space complexes for industrial use or storage. In constructing a large underground room in a mine or for a hydroelectric power project. or the project or excavation will founder in delay and excessive cost. in which relatively large rock deformation accom- panied by new crack growth cannot be ignored even if the rock mass is highly fractured.

sound conditions for the users of the space. It is the flow of stresses tangentially around the underground opening that makes it relatively safer than its surface counterpart. provide surface cuts for foundations of buildings and dams and for quarries attached to specific construction projects. the tengential stress around the opening tends to hold the potentially moving rock blocks in place. The idea will be to initiate self-destruction of the rock through a properly executed initial excavation that triggers the caving process. permanent openings. For example. In the case of cuts and underground openings thatareintended as permanent features. size. these stresses may themselves initiate new cracking. . however. and orientation. or excavations into weak rock. This is the correct approach wherever the mode of failure of the excavation involves the movement of rock blocks. rather than its presence in elevated magnitudes. it is useful to examine briefly all the important failure modes that can actually occur in an excavation. even though mines that exhaust the ore in a few tens of years can one day be abandoned. 1 Temporary excavations that will eventually be backfilled with earth or concrete. block theory will permit the design of excavation shape. or allowed to collapse. In the case of mining openings that are intended to cave. The theory developed in this book applies to all the types of excavations mentioned above. More often. and for underground crushers. as shafts to gain access to a long-life mine. This theory relates to the movement of joint blocks that are liberated by the artificial surfaces to be excavated. which lacks a completely encircling contour to confine rock movements driven by gravity. To appreciate the place of this approach in the total scheme of things. it is usually the absence of continuous compression around the opening. the requirements for safety usually dictate that they be excavated as if they were to be stable. lntroduction Chap. In all these excavations the loosening of rock blocks along the contour of the excavation is to be avoided to achieve safe. In fact. and a system of internal supports (if necessary) that provides stability at minimum cost. There are also many uses for per- manent underground mine openings: for example. But rock falls in the confined space of an underground chamber or tunnel are apt to be more troublesome than even large rock movements into surface cuts. thereby acting as a stabilizing factor. Moreover. but all under- ground excavations undergo localized "overbreak" as the desired dimensions of the opening readjust themselves to geological realities. Surface excavations are less stable than underground openings and experience failures more fre- quently. MODES OF FAITURE Failures of underground excavations are uncommon. In very deep mines. block theory will help choose initial undercuts to maximize the chance of natural rock caving. for hoist machinery. fortunately. or even violent rock bursts. In the mining field it is sometimes preferred to create intentionally unstable excavations so that the costs of rock breakage will be minimized. that permits serious overbreak and failure.

This is likely to be critical when running water is allowed to travel along the surface of altered or weathered rocks. including blocks several cubic meters in volume. and other rocks with disseminated clays (usually smectites). is frequently experienced in these weaker rocks. for example of sulfide-bearing rocks. One form of "failure" of both surface and underground excavations that is not addressed by this book is destruction of the surface by erosion from water. that fell from the roof of a Norwegian tunnel in 1960 (Fig.3(a) depicts the formation of deep gullies by the action of rainwater charging down a face in friable. slow crack growth. 1. shales. (From Brekke. can cause related problems. due to creep. in relatively weak rocks at moderate oreven shallow depth . 1972. roof. such behavior can eventually undermine contiguous rock blocks to permit larger cave-ins.2 Rock fall from a Norwegian tunnel. Occasionally. 1.2. Flatter slopes are more seriously affected than very steep slopes. Loosening and ravelling of shales in underground galleries. with damage to support systems. and floor.Modes of Failure Figure 1. can crush the support and lining structures under the worst conditions. The slide initiated from a tabular mass of altered metamorphic rocks between two seams of swelling clay whose squeeze into the tunnel left the rock free of normal stress and therefore unsupported along its sides. tuffaceous volcanic rocks. Similar results accompany the swelling of active clay minerals in certain claystones. or poorly cemented sediments. As was seen in Fig. problems of active weathering.) consider the slide of some 200 m3 of rock. The inward movement of the walls. or in harder rocks laced by seams or affected by hydrothermal alteration or weathering. or slowly increasing load on the rock. In layered rocks. 1. Another rock failure mechanism not examined in this work is rock "squeeze" depicted in Fig. Figure 1. or gravity alone.3(b).2).

. (g) tension cracking and sliding after loss of toe support.3 Modes of failure of rock excavations: (a) erosion. (f) wedge sliding. (b) squeezing. (e) toppling. (d) buckling. 1 (c) (d) Figure 1. (c) slabbing. lntroduction Chap.

major cave-ins can materialize. 1980.3(d). In the absence of governing geological structure. 1. The interaction of new cracks with preexisting fracture surfaces presents an important. This mode of failure is therefore not discussed here.Modes of Failure (f) (s) Figure 1. as shown in Fig. Excavations in regularly layered rocks present failure modes due to flexure of the layers. but intractable analyt- ical probtem. It would be possible to enlarge the scope of block theory to analyze a situation like that of Fig.4). as shown in Fig. 7. new crack growth will lead to major failures underground only if these cracks intersect preexisting discon- tinuities. 1. A maximum inclination of the tangential stress with the normal to the layers is determined by the angle of friction between the layers (see Goodman.3(c).4. as shown in Fig. When that happens.3 (Continued) and in hard rocks at very great depth.4 if the orientation of the new cracks could be known in advance. Sec. Flexural movements . Therefore. I. This occurs in those regions of the excava- tion surface that make small angles with the rock layers. new cracks tend to develop parallel to the excavated surface. 1. growth of cracks can undermine portions of the excavation and permit local rock failures.

7 and 8)." shown in Fig. 1972. Calif. 1 Figure 1. Thus a serious failure can occur retrogressively.. The main thrust of this book is that modes of failure like that of Kemano tunnel cannot occur if the initiating block is held in place. 1.4 Rock fall in diversion tunnel for Castaic Dam. Cases arise where it is the movement of the first block alone that causes the total failure. Chaps. leading to a style of failure known as "toppling. 1980. What ls discussed is the movement of rock blocks from the surfaces of the excavation.3(f) illustrates the movement of a single rock block from a surface excavation.5 m in diameter suffered a rock fall 20. Such a large mass of rock could cave into the small space of the tunnel section only by retrogressive action.) of a cantilever type occur in surface excavations. while bending of both symmetrical beams and cantilevers occur in undetground openings (see Goodman. the movement of a large wedge in the left abutment of Malpasset arch dam caused that tragic failure (Bernaix. potentially caving mass but only the first block to go. (From Arnold et al.where an unlined tunnel originally 7. Thus we need not analyze a complex. The movement of a first block creates a space into which previously restrained blocks may then advance.3(e). sometimes very quickly. Figure 1. 1966). An important example is afforded by the Kemano tunnel in British Columbia (Cook et al. 1962). For example.000 m3 in volume.. . Introduction Chap. All the modes of failure discussed above are ignored in this book. The cave extended 42m above the original roof of the tunnel.

the procedures determine the most efficient orientation and shape of the initial cuts to assure caving. and other fields of application. civil engineering. tensile stresses were created due to the sliding tendency of a rock wedge along the bedding. in a mining scheme involving intentional rock caving. and changing strength values. The procedures to be presented here demon- strate further how to modify the shape of a gallery. to enhance stability. and almost surely in the design stage. it was new cracking and tensile failures of the rock that permitted the tunnel support failure sketched in Fig. even relatively pure block sliding modes begin to look more complicated by the time the failure has developed because the movement of each block gives birth to new block movement potentials. releasing the wedge.4. an open cut. For example. or the contours of a surface cut. or an underground chamber may be allowable even in the construction stage. Some excavation failure modes are truly hybrid in that the initiation of failure arises from the combination of two or more mecha' nisms acting simultaneously. new rock cracking loci. Although many interrelated factors affect the design of an excavation. in the case of block movements we will be able to show in this book that rational procedures can determine the best orientation of the excavation in order to obtain stability with minimum support cost. shifting loads. The examples above demonstrate that excavation engineering demands attention to the rock structure. the possible applications are many and the problems addressed are significant in mining. Turning a tunnel. Corresponding examples of single block movements with or without new crack growth can be found in underground excavations. ASSUMPTIONS OF BLOCK THEORY The thrust of this book is to produce techniques to specify the critical joint blocks intersecting an excavation. The rock was interbedded friable sandstone and shale. . The engineer can deal with most rock conditions if the geological data are properly incorporated in the design process. Conversely.3(g) suggests a more complex mechanism in which the release of the block is possible only with local rock crushing (at the toe of the lower block) and were release of the lower block triggers tensile cracking and a further block fall from above. heightening the excavation. Many failures of rock excavations develop from the movement of a single rock block previously defined by joint intersections. The rock then cracked. sufficient flexibilities will usually exist to accommodate geological requirements. dipping steeply into the tunnel and striking parallel to the tunnel axis. On the other hand. whose movement into the excavation destroyed the steel supports and triggered a complete collapse of a long section of the tunnel. For example. It applies to rock engineering for excavations in hard rock where the movement of predefined blocks precipitates failure. When the bottom heading was removed. 1. Overbreak occurred in the roof when the top heading was driven and undermined the beds. Even though other modes of failure are ignored.Assumptions of Block Theory Figure 1.

excluding failures with new cracking. no discontinuities wilt terminate within the region of a key block. However. for example. 4. some one direction will have to be taken as representative of the set. In view of the preceding discussion. This means that block deformation and distortion will not be introduced. The implications are that all blocks are completely defined by preexisting joint surfaces so that no new cracking is entailed in the analysis of block movements. . as in Fig. none of which include by themselves more than two joints of any set. which are found through block theory. that locus then being input as an additional discontinuity plane. Introduction Chap. The discontinuities and the excavation surfaces are assumed to be deter- mined as input parameters. this limits the applications to a specific type of failure mode. 1 The problem posed is limited in scope-to find the critical blocks created by intersections of discontinuities in a rock mass excavated along defined sur- faces. that is. Through Monte Carlo simulation techniques. This will not be examined in this book. it is entirely reasonable to treat the discontinuity orienta- tions as if they were precisely determined quantities. 2. 1. on the limbs of folds. 1. it should be possible to examine the influence of variations in these angles and to relate the output results statistically in terms of probabilities. Subsequent examination of the stability of the key blocks. Since the development of frictional resistance along the faces of the key blocks actually entails deformation along the surfaces of the blocks. Blocks defined by the system of joint faces are assumed to be rigid. the latter could be studied using the methods to be developed if the locus of new cracking were defined initially. in which case constitutive relations for the rocks would have to be defined. The principal assumptions follow. Yet this problem is sufficiently difficult that a series of simplifying assump- tions must be adopted in order to gain workable solutions. But this will not be pursued here. it would be a straightforward extension of the theory to permit certain planes to be curved. will then introduce strength properties for the discontinuities.4. it therefore implies accumulation of strain and stress within the blocks. We assume perfect planarity in order to describe block morphology by linear vector equations. but not for all and can be quite wrong for bedding surfaces. All the joint surfaces are assumed to be perfectly planar. This approaches reality for most joints and faults. Deformation of the blocks under limiting surface forces could be examined by coupling block theory to a numerical analysis embracing mechanics. 3. On the other hand. Since the result of the block theory is a list of key-block types. The key-block problem is formulated entirely through geometry and topology. In practice. Joint surfaces will be assumed to extend entirely through the volume of interest. it will be necessary to consider the finiteness ofjointsinattempting to apply the block theory to successively larger excavations. If joint set orientations are actually dispersed about some central tendency.

An example of the output of a block analysis for a tunnel is given in Fig. With the determination of the key-block types. The analysis is three-dimensional. It will be instruc- tive then to compare block theory with the various alternatives. and timiting equilibrium analysis.5 Intersection of a key block and a tunnel. and flexure of layers.Comparison of Block Theory with Other Analytical Approaches 11 In summary. physical model techniques.will be able to analyze the system of joints and other rock discontinuities to find the critical blocks of the rock mass when excavated along defined surfaces. Only block movement modes are to be considered. and discrete element analysis). we will not attempt to assess other modes of failure. Figure 1. GOMPARISON OF BTOCK THEORY WITH OTHER ANATYTICAL APPROAGHES A number of analytical tools are available for engineering calculations involving excavations. These include numerical methods (finite element analysis. we. typical output of analysis using block theory. finite difference analysis. It is assumed that continuum mechanics is second in impor- tance to the calculation and description of key blocks. rock cracking. Most engineering decisions involving rock excavation are conditioned as well by intuitive assessment based on experience or informed judgment.5. Block Theory and Finite Element Analysis By means of block theory. 1. Block theory is new. fn some respects it is more immedi- ately applicable and potent than any of the older approaches. block theory will be developed on the basis of geometric information derived from structural geology and equilibrium calculations using simple statics. . such as erosion. the theory then provides a description of the locations around the excavation where the key block is a potential hazard.

where a two- layer rock beam spanning a symmetrical excavation is modeled. 1. the tunnel section. . Then. [n contrast. Also to be input is the set of mechanical properties for each element of the mesh. Figure 1. But it cannot provide much help in charting the wisest direction for the excavation. 4. I The largest key block. Figure 1. 3. Finite element analysis determines the stresses and. with difficulty. Note the shear deformation of the joint ele- ments over the abutments of the beam and the opening of a gap between the layers in the center of the beam.12 Inttoduction Chap. Finite element analysis must always compute from a specific mesh. In this particular program the input properties to be supplied are the deformability and unit weight properties of the rectangular and triangular elements. generic aspects of the problem posed by a given joint system without calling for a specific joint map. as an optional second stage. defined by the input sets ofjoints. Block theory does not find stresses in or between elements. rather. a list of dangerous or potentially dangerous blocks behind the surface of the excavation. or to analyze it further in hopes that available friction on the faces will hold it safely in place.6(b) shows one part of the output-the deformed mesh-obtained from the displacements of each "nodal point. The latter determines.6 shows the state of stress in the center of each element. by changing the tunnel shape or direction. An example of finite element analysis is shown in Fig. once a model has been set up. block theory analyzes the essential. Finite element analysis can be used parametrically. the theory could be redirected to a new set of input. Finite element analysis determines strains and displacements throughout the model." each corner of an element in this case. The input information required begins with a computing 'omesh" establishing the size and shape of the domain to be studied. is drawn in relation to the tunnel. Alternatively. whereas block theory does not determine strains or displace- ments anywhere. and the stiffness and strength properties of the lines ofjoint elements between the two layers.iately locates danger points and provides an estimate of the support forces needed to avert failure. 2. Block theory can handle both tasks very well. The next step will be either to provide timely support to prevent the movement of this block. as revealed by vector crosses aligned and proportioned to the directions and magnitudes of the principal stresses. and the tunneling direction. Generic studies can be made only if numerous meshes are generated. with predefined directions and spacings of joints. which in this symmetrical example defines half of the roof. 1. to find a suitable shape for an excavation. Block theory immed.6. the theory can be applied specifically to actual joint locations (if the data are available). We have seen in these examples of block theory and finite element analysis that there are fundamental differences. these can be manipulated to find regions of potential danger. for example.

(c) stress field. (b) deformed mesh.+ | +++ + I I rl|l (cr Figure 1.Comparison of Block Theory with Other Analytical Approaches 13 (al (b) tt +f +-y + t*+ + + i ii.6 Results of a finite element anatysis of an excavation in jointed rock: (a) initial mesh.) . (From Hittinger. 1978.

In contrast. and Voegele (1978). block theory computations are far less costly than finite element runs. 1. both because it is independent of the precise joint map in the section of interest. In the distinct element programs pioneered by Cundall (1971). A finite difference or finite element analysis of the system can then trace the rotations and displacements of the block system as conditioned by the load/deformation relations adopted for . incorporating precise loca- tions of all joints. The graphical techniques required to apply almost all aspects of the theory will be demonstrated in this book. relatively large two-dimensional block systems are calculated by integrating finite difference approximation of the equations of motion for each block. Since these programs are relatively small.) Distinct element analysis is an instructive tool for excavation engineering in that it permits analysis of large block movements in geologically complex sections having many joint blocks. Block theory is better equipped to help choose the direction and shape of an excavation. these programs are suited to micro- computers of the type now available to most engineers. It is restricted to two dimensions.the joints. for example stereo- graphic projection. . As the computation begins. This infor- mation had to be input as well as the friction properties of each joint and the unit weight of all blocks. speciflc applications can be freed of tedium by making use of a computer. finite element analysis is a larger computation than block theory and will always require a computer.7(a) shows the initial positions of the blocks. Block Theory and Distinct Element Analysis The distinct (or discrete) element method is a numerical model approach with reduced degrees of freedom compared with finite element analysis. As with flnite element analysis. In contrast.14 Introduction Chap. On the other hand. however. (The reason for this can be demonstrated using block theory. The analysis can be performed with a microcomputer and displayed interactively. Figure 1. By removing deformational modes from blocks outlined by joint elements. with changing boundary forces calculated at each time step from the changing block interactions. As noted earlier. block theory can be applied entirely manually with graphical methods.7(b). only rigid-body modes remain. a second model with joint AB rotated clockwise to a more nearly vertical position produced instability and collapse. An early stage of output is shown in Fig. In general. and because it is three-dimensional. block theory does not offer an analysis with large deformations. The model achieved stability through arching as the program continued to run. 1 5. In contrast to finite element analysis. it is still necessary to compute from a predetermined mesh. the blocks displace and rotate under gravity and the deformed mesh can be followed through large deforma- tions. block theory does not require premapping of the joints and it is fully three-dimensional. On the other hand. unless very large computers are used.

have been used to relate excavation directions with joint set orientations.7 Distinct element analysis of a tunnel in blocky rock by Voegele (1978): (a) input information. Experience offers no alternative to rational procedure when designing an excavation of unprecedented shape. if any. which is tailored precisely to the three-dimensionality of the problem. can address the siting of excavations in better focus than can intuition. Block Theory versus Engineering Judgment Engineers have been siting excavations in jointed rock for a very long time longer than the period in which numerical tools have been available. (b) output after an early stage of deformation. it is not clear what procedures.5" {bl Figure 1. except for the general rule that major rock walls should not be cut parallel to the strike of a major joint set.Comparison of Block Theory with Other Analytical Approaches 6 = 26. . experience. Block theory. The excavation siting problem is truly three-dimensional. It is noteworthy that relatively few excavation engineering experiences haYe been well documented in the literature so that a newcomer cannot easily acquire such experience from study alone. -far Presumably intuition. hopefully. and judgment were called and. some specific information about the directions and properties of the major joint sets. or function. size. Further.

1977.g.E Londe's analysis of stability of a dam abutment.. the value of analyzing the limiting equi- librium condition is well appreciated. Bishop's modified method of slices.) to determine the critical condition when the soil is just about to pass from a state of stability into one of instability. Introduction Chap. Referring to (". 1 Block Theory and Limit Equilibrium Analysis In the field of soil mechanics.) . (__: fr\ lo €o I tl 2 /60 '7O U3 Wl U1 l%l Figure l. Methods of limiting analysis for slopes and foundations are available (e. etc. Morgen- stern and Price's method. (From Londe and Tardieu.

9(b) shows an even more complex. In Fig. Such models permit structural engineers to assess and design for rock weaknesses due to discontinuities. and can be quite sophisticated. Each face plane l. however. provide the only three-dimensional alternative to block theory with respect to practical rock problems.1. and Ur) in the diagram produces a point R (or R') whose position determines two strength parameters. they are not convenient for under- ground excavations. volumes. 1. 1. For example. and p. Methods for assessing limiting equilibrium also enjoy use in rock mechanics. they are too cumbersome and costly for everyday problems. Limit equilibrium analyses for rock wedges are discussed in detail in Chapter 9. face areas.9 gives examples of some impressive physical models tested in the laboratory ISMES in Bergamo. because three-dimensional physical models have great visual impact.Comparison of Block Theory with Other Analytical Approaches 17 the limiting state saves the trouble and uncertainty associated with trying to define the actual state of stress for a specific set of conditions. We will extend the previously available analytical methods to permit stability analysis of blocks with any number of faces and any shape. We wish to point out. (Jr. . No three-dimensional models have yet incorporated block shapes more complex than rectangular prisms.8 shows the results of Londe's equi- librium analysis for a tetrahedral rock wedge within the abutment of an arch dam. a necessary prereq- uisite since it will allow you to determine which block to analyze. and other factors that are required for further analysis. three-dimensional block model used to design a reinforcing structure to stiffen the abutments of a high arch dam abutting against limestone with inclined joint sets. Figure 1. Nevertheless. rather. or 3 is potentially subjected to a water force equal to a given propor- tion (U) of the head of the reservoir acting over the entire plane. Although such models are useful in practice. Block theory is not a substitute for the limiting equilibrium analysis but. Fig. It will also evaluate the blocks' dimensions. and embracing full similitude in all important physical quantities. Thus the inherent anisotropy of the rock mass is not accurately represented except in those cases where the jointing is actually prismatic. Italy. that such an analysis can be run only after a particular tetrahedral block has been singled out. they are here to stay. The wedge has three joint faces initially in contact with rock. in which the region to be observed is hidden from view. Furthermore. 2. Figure 1. that determine the degree of safety of the wedge. . It is beyond our present purpose to describe the solution or these parameters at this juncture. This model was made with carefully fitted prismatic blocks. Entering appropriate values for the three water force coefficients (Ut . Block Theory and Physical Models Three-dimensional physical models of rock excavations containing joints.9(a) the rupture of an arch dam model bearing on an abutment in layered rock is shown to follow from interlayer sliding in the rock.

(From Fumagalli. 1978. (b) a dam on steeply inclined layers.9 Physical models of blocky rock foundations after testing to failure: (a) a dam on gently inclined layers. (a) (b) Figure 1.) .

The Key Block System THE KEV BTOCK SYSTEM The objective of block theory. 1977). Figure 1. is to find and describe the most critical rock blocks around an excayation.11(b) shows key . or pins (see Barton. Some of these will not be able to move into the free space of the excavation. as stated earlier. 1. other blocks that were previously restrained will be liberated. when the excavation is made.10 because the joint blocks around an excavation are not perfectly similar in shape. 1. This model is more appropriate for excava- tions than the Roman Voissoir arch of Fig. the arch will function.10 Sketch of part of a Roman aqueduct in Spain. with a shape different than the rest. As long as it is kept in place. is restrained by bolts. then. or orientation. One last block.10 shows two foundation arches of a Roman aqueduct in Spain that stands and supports load without bolts. Another type of masonry arch is sketched in Fig. The intersections of numerous joint sets create blocks of irregular shape and size in the body of the rock mass .11(a). A few blocks are immediately in a position to move. either by virtue of their shape. (From Barton. and as soon as they have done so.) masonry arch every block is key because the loss of any one of them would cause the entire structure to collapse. Figure 1. 1977. many new blocks are formed with the added surfaces. or because they are prevented from moving by others. and a moderate force may be sufficient to achieve that.In this Figure 1. size.

. and so on. Rock slopes of surface excavations show. blocks around an underground opening. I Figure 1. Loss of the shaded blocks (1) would permit movement of blocks (2). then (3). without dynamic numerical modeling. (b) an underground chamber. but it is definite that survival of block (1) wilt curtail movement of any higher-numbered block. similarly. (c) a surface cut. exactly what will be the sequence of failure afterthe removal of block (1).11(c). dependence on a few critically located blocks. as in Fig. destroying the chamber. Introduction Chap. 1.11 Key blocks in: (a) an arch. It may not be certain. (d) and (e) dam foundations.

then (2). in (b) the tunnel azimuth is 15'.12 Influence of the direction of a tunnel on the extent and nature of rock falls. set (3) 90" and 15o. Even afterward.The Key Block System 21 Figure 1. The dips and dip directions of the joints in the rock are as follows: set (1) 60'and 285'. In (a). translating or rotating indi- vidually. . The block drawn is the largest of its type that can fit into the space of the excavation or natural valley in which the structure is located. The block shown is to be restrained from uplift of water and earthquake forces by the downward forces of cables that are anchored below the key block. All these examples try to show in two dimensions what is usually under- standable only in terms of three. Plane p below the dam would appear to be a possible sliding surface to be calculated for the equilibrium of the dam.12(a). 1. But the rock above p cannot move as long as block (1) remains in place. The joint directions and spacings of the rock mass are given in Fig. If the tunnel is oriented in the direction of azimuth 105". showing the key block of a foundation and its relation to the design of anchors to hold the structure down. then (3). the large foundation mass above plane p could not move without lifting the dam. set (2) 48" and 105'.L2. Figure 1. with first (1). They are therefore greatly simplified. Arectangular tunnel6 m wide by 3. the rock will not cave. 1. except for (a) Figure 1. Another example where the three-dimensional character of the key blocks is essential is shown in Fig. the tunnel azimuth is 105". and so on. but it could be destroyed byretrogressive action.11(d) shows key blocks in a dam foundation.5 m high is intended to induce caving of the overlying rock to permit its mining.11(e) is a more realistic example.

-f- __L Figute 1. Introduction Chap. it caves as shown in Fig.13 A wedge that is daylighted when a portal is created at the toe of a slope.12(b). This block -a. If the tunnel is turned to azimuth l5'.12 (Continued) the single shaded block above the roof. 1 (b) Figure 1. 1.=" .13 shows the intersection of a key block with the excavation created at a portal. The portals of tunnels are another environment that requires three-dimensional theory for solution. Figure 1.- . .

Movements of blocks like those sketched in these examples have produced numerous important failures. and are responsible for routine overbreak and rock falls accounting for great cost. . including complex three-dimensional openings at tunnel intersections. nor into the free space above the valley side. or near failures. bifurcations. portals. The methods of block theory can be applied to prevent these types of problems if the requisite information about the joint sets can be obtained. including vector analysis and stereo- graphic projection. enlargements. The first chapters present the methods we will use to treat these problems. The particular applications to be discussed in later chapters are: foundations and surface and underground excavations. and so on. but it can slide where one meets the other.The Key Block System 23 cannot slide into the tunnel.

The dip direction is defined by horizontal angle p from y toward x. using the vector equations to produce and display graphical solutions on the screen of a microcomputer. applied forces. Solution of a set of vector equations with a computer is facilitated by entering them in coordinate form." An inclined plane.L explains these terms and their relation to the geological quantities known as "strike" and "dip. The fundamental information required by block theory is the description of each joint plane's orientation. In practice. The methods of vector analysis provide relatively simple formulations for all the quantities relating to block morphology. The next chapter develops an alternative graphical solution procedure. intersects the horizontal xy plane along the "strike [ine" and plunges most steeply in the "dip direction. from which the coordinate equation are generated. including the volume of a joint block. the area of each of its faces. More compact vector formulations." whose average orientations are described by two parameters: the dip and the dip direction. In this chapter we develop vector equations that permit computational solution for the basic problems of block theory." which is perpendicular to the strike. Throughout 24 . and friction. we often combine both methods. are written here as well. and support reactions. some mathematical preliminaries will be helpful. according to a Cartesian basis. The joints are collected in a modest number of joint "sets. and the positions and attitudes of its faces and edges. Figure 2. inertia. the positions of its vertexes. ruled on the figure. The use of vectors also permits kine- matic and static equilibrium analysis of key blocks under self-weight.F chapter z Description of Block Goornotry and Stalcility LJsing Vector Methods Before proceeding with the fundamentals of block theory.

Y. If lines in the upper hemisphere are preferred. the opposite to the dip vector is defined. UsuallY. The amount of dip. rising at a in direction f + tgO'. w€ will select the upward normal. dip angle from horizontal. EOUATIONS OF LINES AND PLANES The general case with a plane or line not passing through the origin will be treated first. f. corresponding to the choice of a > 0.with direction p from / (towards x) and inclination c from z (vertical). Either an upper hemisphere normal or its opposite in the lower hemisphere could Ue required in a particular analysis. as shown in Fig. Later we examine the important special case with all lines and planes containing the origin.Equations of Lines and Planes Figure 2. In this manner. 2. dip direction (clockwise from north)' this book. we adopt the convention that y is north and x is east.1Terms describing the attitude of an inclined plane: a. with z up. Let om be the direction of dip of a plane passing through o and dipping a with horizontal. and z' . the angle B is the same as the compass bearing of the dip direction as conventionally given in geological reports. Then the upward-directed normal to the plane is the tiii on. An alternative to expressing the orientation of a plane by its dip vector is to state angles s and p tor its normal. or simply "the dip. Coordinates are represented by values of the intercepts X.2 (with a measured from vertical). y. and Z along the basis directions x." is measured by vertical angle (a) between the dip direction and the trace of the joint in a horizontal plane.

Z). (2.. Substituting x:(X.3).2) and Xr : (X1. 2. z) (Fis.Z) xo : (Xo.1) can be transformed to coordinate form by replacing each radius vector by the coordinates of its tip. Y. Yo. Equation of a line. Yr. Description of Block Geometry and Stability Using Vector Methods Chap. Z) is defined by the set of points along the tips of a family of radius vectors x such that X:Xo*txr Q.T) where xo is the radius vector from the origin to point (xo. be a "radius vector" from the origin to point (X.1) generates three parametric equations that are its coor- dinate equivalent: X:Xo*tX. normal of joint. dip angle. Yo. A line in direction x..2 Coordinate system and direction cosines of a normal: z. a. f. Z) in vector equation (2. 2 [- x east Figure 2. Zo) (2. dip direction (clockwise from north). The vector equation (2. projection of n to plane OXY.Y.3) Z:Zo*tZr . Y:Yo*tY. z. through point (Xo. Let x. The parameter / takes any negative or positive real value. Yo.

B.4) where D is constant.the values of the normal's coordinates are I : sin usin p B:sinacosp (2.6) As shown in Fig. D (2. The equation of a plane. . 2.* Equation (2.2. C) to yield AX+BY+CZ:D (2.4) can be converted to coordinate form by the following substitutions: x: (x.s) and fio : (A. Y. Let fio be the unit vector directed normal to plane P and x be a radius vector from the origin to any point in plane P.Equations of Lines and Planes x0 Figure 2.7) C: COS fl *The symbol ^ over a lowercase letter will always signify that the letter represents a (a unit vector direction). D is the length of a perpendicular from the origin to the plane.3 Equation of a straight line. As shown in Fig. 2. fio. The plane P is defined as the set of tips of radius vectors x such that x.4.z) (2.

which is different from unity. A point like c of Fig..4 Equation of a plane.e) and Z:Zo*toZr The line of intersection of two planes. 2 The intersection of a plane and a line. Z) gpven by X: Xo * tox. Since the line of intersection is contained in each plane. Zo) be a point on the line that has the direction of a radius vector to point (xr. (2. and firbe unit normals to planes p. and the radius vector from the origin to the point of intersection of the line and the plane has its tip at point (X. . Substituting the values for x. and pr. Let(Xo. it will be noted by Ir". and each plane contains only the lines perpendicular to its normal. Yr. and frr. and Pris parallel to Ire : fit x fiz (2.6) and solving determines I as t:t--to 5 D-(AXo+BYotCZo) .3) in the equation of the plane (2. Let fi.2.3) and(2. is perpendicular to both fi. If only the direction of the line of intersection is needed.5) with line of intersection f . Y:Yo*toY. andz ftom (2. Consider planes p. Y. The intersection of two joint planes creates an edge of a joint block. and p" (Fig. Therefore.4.8) Figure 2. can be described by solving equations Q.r.2.6) simultaneously. A line that is perpendicular to two other lines is generated by the vector ctoss product. then I12 has magnitude l(f r x fiz) l. the line of intersection of P. then f. Yo. zr). - (2.[" Description of Block Geometry and stability using Vector Methods chap. y.10)* *If the planes are not perpendicular to each other. where a line pierces a plane.

6 shows a polyhedral block. let fi. lx' Y'll. and 2 be unit vectors parallel to the coordinate axes. Zr) and let ft.which solves the set ArX*BrY+C&. vertex A. Y2 z'l* .X"Yt)l (2'12) A corner of a btock.Z). : (X r. . Figure 2.Y. zrl' - Irz: lx" Yrl" In coordinate form. z'l* .(XrY. . Y r. anilPr. j. lrr: l(YrZ. For example. . Then since x 2 it | xr Yr Zrl I fi.Dl ArX*BrY+CzZ:Dz (2.11) x2 Yz Zrl I Yt Zrl^ lX.l*.Yr. To convert this to coordinate form. defined by the intersection of planes Pr. x fir: (2.YrZr).t3) ArX*BrY+CiZ:Dz .5 Line of intersection of two planes. The coor- dinates of its corners (vertexes) are each simultaneous solutions for the equations of three intersecting planes.Equations of Lines and Planes Figure 2. is determined by point (X.XrZr). Pz.(XrZ. then. Zr) and frz : (Xz.

Let the equation of plane p be AX+BY+CZ:D where (A.e. consider plane p of Fig. a point x : (x. 2 D Figwe> 0). is upward (i. B. 30 Description of Block Geometry and stability using vector Methods chap. The description of a half-space. in coordinate form.z1 will be said to belong to the lower half-space of p ii fi-. AX+BY+CZ<D (2. If normal fi.7.r6) or. AX+BY+CZ>D (2.6 Corners of a block. 2. a point x: 1x. i. that is. point C1 is in the lower half-space of P.17) .t4) similarly. C) defines the coordinates of the tip of the radius vector fio per pendicular to planeP. x> D Q. z) will be said to belong to the upper half-space ofPif fi-.x<D Q. A point like C2 is in its upper half-space. in coordinate form. it is above plane p. y.. Determining whether a point is situaied above or below a plane is a cornerstone of block theory.

ao.7 Half-spaces determined by a plane' DESCRIPTION OF A BLOGK We are now in a position to determine all the relevant features of a block-the numbers.h. and its volume. as shown in Fig.and ar.g.8(a). with corners 0t. and areas of its faces.8(b). * divide it into two equivolume triangular prisms by . 2. Each of these. in turn.031o4s o5e or. locations.8. . *. 2. the locations of its corners. \\ \ Figure 2. area ata2atai of Fig. as shown in Fig. 2. First. This yields tetrahedra with corners ara2a3q4)azasa4as. Consider the paral- lelepiped drawn in the upper half of Fig. 31 Description of a Block \{f \\'\ Dz\ \ \ \\'."tting along plane a2qjasa6. Since the volume of the parallelepiped is the area (S) of its base [e. 2. The Volume of a Tetrahedral Block A four-sided block can be thought of as one part of the division of a paralletepiped into six equal volumes.8(a)] multiplied by its .andarana. Q2. can be divided into three equivolume tetrahedra.

Volume = * hS . 2 la1 a2asaTaaa5 a6 ag) Volume = hS a4 a6 -A (a1 a2a3aaasa5) .'/ \. (b) division of each prism into three tetrahedra. auau ar) Votume = jtrs (a) Figure 2.I92 Description of Block Geometry and stability using vector Methods chap.8 Subdivision of a parallelepiped into six equal tetrahedral volumes: (a) subdivision into two triangular prisms. .tt / tr a5 (araa a.

ara3 aa) Volume = t trS Volume = 5 (b) Figure 2.Description of a Block 33 (a1 a2 a3aaa5a6l votume = * sh (a.8 (Continued) ..

and c. a: (X.1e) and c: (Xn .Yz .Oxc) (2. b. Yr-Y. Yr-Y. .I 34 Description of Block Geometry and Stability Using Vector Methods Chap.Zr) b: (X. ar. are the three edge vectors radiating from where. any vertex of the tetrahedron. Xr. the tetrahedral volume is expressed in coordinate form by Ix"-x. 23} Flgure 2. b.Zs Zr) (2. YrrZ.2 height (h). . y2. it follows that each tetrahedron must have volume tSU in vector form. as shown in Fig. .18). za) bXc a1 (x1.I9) into (2.yr. Zr-2. 2. ct2r 43' and a4' and taking ar as the vertex from which radiate vectors a.oo:*a. Y1.o"n. Y4 . Zr-2. Zn-2. z1) a3 {x3. ya. l1 Xt Yt 'rl v : +lr x' Y2 zrl "ll xs Y3 zrl Q. Xr. Zrorr. (2.18) t.9 Vector names for edges of a tetrahedron. y3.20) lxn-xr Yn-Y. Xr. z2l ?4 lx4. Zr) Substituting (2.9.2r) l1 x4 Y4 znl . Alternatively. v-*lxr-x. Zn . and c. Letting the four corners of the tetrahedron be a2 (x2.Ys * -Yr. *In the case of a vertical plane. denoted [f. For each plane i.22b) if the block is defined with Z-. The procedure for making this selection. in general. C. only eight will be real. i. The number of corners calculated in step 2 equals the number of combinations of n objects taken 3 at a time (Cj). Calculate the coordinates of all possible block corners.. is always positive. For the present. Consider a three-dimensional block with n faces formed by portions of n planes. Edges. meaning that of the two possible directions for the normal the upward one is selected.Description of a Block The * B^Yur.o satisfy A^X.13). Examine every possible corner C.e. the coefficients A. must be in turn and retain it as a candidate-real-corner if its cordinates Xro. Consider face m. or A^X1o * B^Yrtr * C*Z. The intersection of one or the other of these half-spaces of each plane (i: I to n) determines the dimensions and morphology of the block. 3. Br. most of which. In later chapters we show precisely how to choose which of the many combinations of Z's and U's will define the critical blocks. using sample field data will be presented later. which equals nr. i: 1 to n. is presented in steps 3 and 4. and a lower half-space denoted L.6) there are then 20 possible corners. amount- ing to simultaneous solution of n inequalities.e.. n ..The dip angle o is always between 0 and 90o.l(n . and C. 2. a block may be created by Lr.... and Dr. + C^Zt*) D* (2.. 1. For a parallelepiped (i. and Z. Our procedure for calcu- lating the volume of any such block is to subdivide it into tetrahedra and then make use of (2. Ur..20). . and k. and Lu. We must now determine which of the corners actually belongs to the block (i. Y. we assume this to be given. and D. A corner Cqp is calculated as the point of intersection of the three planes i.22a) if the block is defined with U^. An example calculation of A. Lr. as described by (2. Cr. are calculated from the dip and dip direction of plane f using (2. will have more than four faces.7).3)! 3!).t* I D^ (2.* The coefficient D. and C. B. For example. Each plane (i) divides the whole space into an upper half-space. which are real). (Jr.. Ln.. the normal is determined as positive in one direction or the other automatically according to the value input for the dip direction angle (f). and Gorners of a Polyhedral Block with tt Faces The intersection of joint planes creates blocks with various shapes. determine the constants A.

labeled Crr. Umeans the half-plane above a line andLmeans the half-plane below a line. and so on (the order of the indexes has no significance). crr. through pr.(Jo. The real corners are those candidates that survive step 3 for every single face.. p. a two-dimensional example will be helpful..e.t) D.t * BtY. At this point. . Lz.the real vertexes (Cri) must all satisfy ArX. half-plane (J1. (Jr. m : ! to n). Description of Block Geometry and stability Using vector Methods chap. The real corners of block (J. Crr. This step eliminates Gc.andU. [Jn. are thus identified as Ca1. and so on. and crr. cnr. Repeat step 3 for every face in turn (i. Ls.10 Real corners of a given polygon. ul L2 L3 U4 U5 x Figure 2. next eliminates Crr.. Now consider one specific block i ur. first.10 shows the polygons created by five lines. (In this example. There are c?:r==#16r: ro intersections of these lines. Considering half-plane L. Figure 2.Lr. 2 4.) Considering.Lz. crr.

Cr"n) and II(Cnr.Face Pn is already a triangle-(Ct . from decomposition in triangles. m.12 gives an example of the procedures described in steps 7 and 8. For example. subdivide each of the other (n . Now subdivide face m into triangles by selecting one corner and connecting it in turn with the end points of each edge of face m.andft. For example. (ararao). The choice of corner is arbitrary and only one corner is to be selected.t2(a). 2. Faces P" and Pn remain. the intersections of the five lines produced polygons of three. into tetrahedra. The corners of face m are the subset of the t real corners of the polyhedron that have m as one of its indexes. Next.. blocks created by intersections of n planes may have from four to n faces. connect the corners of every triangle for all (n . j.1. The five- .Description of a Block 37 We have now succeeded in identifying the coordinates of all the corners C. (The edges of polygon m are the subset of all edges.11 illustrates the procedure described above. one edge is the line connecting corners C. Figure 2. and D ut.oisthe point of intersection of face planes i. and then into tetrahedra by subdivision of the polygonal bases into triangles. found in step 6. 6.P3. the sum of volumes of which is the volume of the polyhedron.3) faces of the block into triangles as follows. as the apex. and k) with the apex. In three dimensions. Finally. polygon (arararanssae) is divided into triangles (ararar). we will find all the real faces of the polyhedron.3) faces (excluding faces i. It divides into two triangles : I (Cs35. Each face. The next steps will divide the polyhedron first into polygonal pyramids. face m (of plane m) is the triangular region between corners Cr-r. that have m as one of the common indexes. Considering again the two-dimensional example of Fig. Cro*. cotnet C. four. (aranar).') of planes i and i. and C-r5. 2. and P. 2.i. First. we arbitrarily select corner C1r.12(b). the five-sided block shown in Fig.n. 5. A real edge is a line between a pair of real vertexes Clyp that have two common indexes. is in general a polygon with t corners. and (ararau).. C. This excludes faces P1. 7. A real face is defined by any subset of three or more real vertexes (block corners) that have a com- mon index. and five faces. Choose one real corner C11p as an apex (the summit of a polygonal pyramid). Crrr. Determine all edges of the block. Crrn).) Figure 2. Excluding these three. We will subdivide.10. This line is parallel to the line of intersection (I'. The former is shown in Fig. By choosing corner al as the vertex of all triangles. Czto). This creates the set of tetrahedra..oof apolyhedral block.. Crro. Determine which faces belong to the given block. 8. Cr2a...

az. and areas of each face of a general n-faced block can be computed using the procedures described in the preceding section. sided block is split into three tetrahedra by connecting these three triangles with apex Crrr.These tetrahedral volumes are shown in Fig. 2 Areaa'a6auai=1bXcl a1 a2 a3 a4 a5 a6 a1 a2 a3 a1 a3 a4 a1 A1 85 a1 O5 E6 Figure 2. and (e). The Faces of a Polyhedral Block The corners. Each polygon is then divided into triangles by the procedure of Fig. 2. (d). edges.IZ(c). A poly- gonal face is defined as the planar region between all corners (C. Consider a triangle with corners (ar.38 Description of Block Geometry and stability using Vector Methods chap.rr) that share any one index. 2.a3) and sides L: atazand .11 Subdivision of a polygon into triangles.11.

Yl 22 zr-2. Y r.Zs . .Zt) giving .xrl' (lv A: -lrll" . . Z. The area of the triangle is U: +la x bl (2. a2 : (Xr. .trl' * z.Yt Z. .z. . Face 2 n q / \ts r/rr ilage lz Edge 24 cnq (b) Czss Cr ss Figure 2.lxr-x| Y2- -lxr-xL Y3..24) .Yr.12 Subdivision of polyhedron into tetrahedra. xr-xrl IIY' . b : erar. . Zt): a : (X. Z r).Xr. : (Xr.'.Xt.Y.Yr. Yr. Y. Yt. . and os : (Xt. x.Zrl (2.23) In vector form with a. Zr).Zr) and b : (X.''']["' . .

(b) the angle between two planes. Ilr : (Xb Yb Z) and [2 : (Xz. yz. Zz) The angle (a) between nl and n. in space.13 Angles between lines and planes: (a) the orthographic projection of a line on a plane.2s) \ /n' \1 \6 / \+L/ \/ (b) Flgure 2. 2 (The formula can be simplified by choosing new coordinate axes x'. !'. and z' uirth z' perpendicular to the plane of the face. or between a line and a plane will be required routinely in computing the sliding resistance of blocks.+ lnrllnzl Q.-=!=.40 Desctiption of Block Geometry and Stability Using Vector Methods Chap. . and n. between planes. Consider two intersecting vectors n. The angle between lines.) AilGLES IN SPACE The angle between lines. is given by cos d .

Recall that a particular block is created by the interesection of upper or lower half-spaces corresponding to each of its faces. But it is appropriate at this point to describe the block pyramid using the formulas established earlier in this chapter. The equations of a plane. that is. The superscript (0) signifies that the plane in question has been shifted to pass through (0. a block bounded by z nonparallel surfaces.27) The angle between a line and a plane.3).e.Zr): (0. be a vector inclined with respect to plane P2. 2.13(a)].2: XrX. The set of shifted half-spaces UIUEU\L\LZ will create a pyramid-the "block pyramid"-with apex at the origin.I4. a block might be given by UrUr(JrLoLr.13(b)'the angle d between two planes P.6).31) Z: tZt Faces through the origin. and nr. COS d : fir. 2. xo : (Xo. are X:/X1 (2. (2. as shown in Fig. Let n. As shown in Fig.0)]. Any plane (i) of the block pyramid will include the origin. is the perpendicular distance from the origin to the plane.0. Lines through the origin. whose normal is nr.. : xg2 + Y\Yz +JA. the equations of the edges. Now let each half- space be shifted so that its surface passes through the origin. : fi. and its line of projection in P" is the complement of the angle d between n. 2. and a2 : fi'.4) and (2. * ItrY" -f Hence Dt :0.1) and (2. The angle between a line and a plane is defined in terms of the angle between the line and its orthographic projection in the plane [Fig.. (2.The Block Pyramid or' in coordinattj. simplify to . The angle y between n.0. d:{(n. (unit vectors). T:90-6 (2.28) The angle between two planes.30) X: tXr Y: tYr Q.26) " If (2. All the edges of the block pyramid are lines passing through the origin [i. cosd:ffiffi :. Therefore. since D. and nr. and P2 is the angle between their normals n. For example. obtained from (2. 0). The importance of this construction will become apparent in Chapter 4.2e) THE BTOCK PYRAMID Consider a real block formed with one each of n different faces. (2.

Description of Ll and Ul.: 1 to n) is at : (Au Br. / | '05 . Every pair of indexes of F0 (i and j * i) defines a potential edge vector Irr ItJ: n. Description of Block Geometry and Stability Using Vector Methods Ghap.X t B. the upper half-space (U. o.14 Block pyramid: U?UgUgLlLg. its lower half-space (^Ll) is determined by (2.. Ct) (2.3e) .35) Similarly. o) Figure 2.33) where tt : (An 8.. C) is normal to joint plane f.Y*C&>0 (2..X*B[+C&<o (2.X*B.38) I-et Fo'F| . When plane (i) cuts through the origin.. The normal to plane P.0). (i. X nJ Q. es4.0) of plane f is determined by (2.36) A.// \/ (o.ul .0 orz. 2 /\ I I I^ P.. F9 be the block pyramid corresponding to a particular set of upper and lower half-spaces (U. X.16) as trr'X)0 Q.37) Edges of the block pyramid.34) or A.14) as n.trf:0 (2.Y * C1Z: O (2.x<0 (2.32) and A.

i) since each pyramid face lies between two edges that share a common index. In fact.fy.) . Fl. Representation of a force by a vector. Zrt) (2. .Z) (2. ltr: ry X nr .Et. frr : (Xr1.40) With n block pyramid faces (n planes).) Figure 2. A series of intersecting forces F.42) arc real edges of the block pyramid. We make use of this in the next chapter. . The sequence of these edges around the pyramid are determined by the numerical sequence of indexes (r.Irr. friction. support forces.. Werepresent both the magnitude and direction of a force F by the symbol F. that satisfy all n simultaneous equations of (2.-ltt (2. for each pyramid face (m)(*: 1to n).. and cohesion. . For example. F: (X.J <0 when F?" : L2' (2. Fn can be replaced by a resultant R: R:iF. . Fg.l4 the edges in order areIrr.43) The magnitude of F is lF l : (X' + Y2 + 22)t/2 (2. in Fig.2. To be the edge of block pyramid F?. There are no more than n such solutions.t * B^Yu + C-Z. .a2a) A^X. water pressure.4s) + Yz * Zzyrz The resultant of two or more forces. .. inertia forces.Equations for Forces 43 According to the rule for the cross product. determined as follows. EOUATIONS FOR FORCES Vector analysis facilitates the analysis of block stability under self-weight. and I23. .M) and the direction of F is given by z (2.I5(a) shows a graphical solution for two-dimensional summations. A*X. Its components are its coordinate values..:(.1. (It is always to use the two-dimensional solution for two intersecting possible forces by transforming the basis directions to coincide with the plane common to the directions of the two forces.42b) The intersection vectors f. Y. . Irn. Fz. the total number of possible edges equals 2C3:n2-n.i*.t + C-Ztt>0 when FI: Ul" Q. a specific block pyramid has fewer edges.4t) must satisfy.t * B^Y.. Iur. y.

2 Figure 2.48) where f. is the friction angle for sliding in direction ^f on face i. Let B signify a potentially sliding rock block and suppose that {. . .. fl. 2. are the magni- tudes of the normal reaction forces from each sliding face P. Fr. . .46) j=1 - E (X.. Then the direction of all friction forces is -.47) i=l Friction forces. Then the resultant friction force is Rr : -E (t.15(b). If a system of n forces F. are inequilibrium. . their resultant R has zero magnitude [see Fig. Friction provides a resisting force that opposes the direction of motion or of incipient motion. . of . Therefore. i : l. We will call the latter the "sliding direction" (^f). If the weight of a block is W.. The equilibrium of forces.r : -(kW)A Q.) : n 0 (2. tan f.. . Gravity acts remotely and its force is proportional to the mass. Y. . ft) equilibrium under forces.Inertia forces act in a direction opposite to an applied acceleration and are also pro- portional to mass. Gravity and other body forces.8. n \-F:0 L/ \'t Q.. . F.M Description of Block Geometry and stability using vector Methods chap. Its direction is vertically downward (-2).i.15 Forces as vectors: (a) the resultant of forces. i : !.. Z. . the inertia force of the block that is accelerated by (kg)d is F. .4e) . fl. .)^i (2.

2. The procedure for calculating the area of any face of a polyhedral block was given earlier in the chapter. Water forces and cohesive forces. structural pull. Suppose that P. the direction of initial block motion coincides with the direction (l) of the resultant force (R) acting on the block . : P. In both cases. are the faces of a polyhedral block. is the integral of water pressure acting over face i.S. or inertia force may pop up a block as shown in Fig. with other known forces produces a circular cone in space that contains all possible resultants. the total force is computed with the known area of the face.r.ft. and inward unit normal.i:f Q. The integration of water (Ft-tl acting over the face of a block produces a water force in the pressure direction of the inward normal to the block. it may fall out under gravity alone.16.Ai (2. Water pressures in rock produced by hydraulic structures tend to vary with time. Ltft@ or fallout occurs when a block loses initial rockirock contact on all faces to advance toward free space.. The action of water pressures. In many cases it is sufficiently accurate to substitute S. of all water forces is I* : j=l n E S.51) where P.If the block exists in the roof. fi.16 Lifting. Lifting. (2. Then the resultant r. If cohesion is constant over a face. each with area A. the summation of F. Cohesion (Ft-'lproduces additional resistance to motion. GOMPUTATION OF THE STIDING DIRECTION The direction of incipient motion of a block is determined by the mode of failure.Computation of the Sliding Direction where g is the magnitude of the acceleration of gravity. If the direction of acceleration (r?) is uncertain. is the water pressure acting at the centroid of face i.50) where . Sliding may occur on any face individually or on two nonparallel faces simultaneously along their line of intersection. .s2) Figure 2.

\['l \\l Figure 2. Description of Block Geometry and Stability Using Vector Methods Chap.xr)xt." The double use of the cross product is justified in Fig... the sliding direction ^i is slll (fi.17 Single-face sliding. A block may tend to slide along a single one of its faces.. and n.I9). 2. 2. The projection of r is along the line of intersection of plane i and a plane common to r and fi.//l s h=fiixR Ftgure 2. Sliding in two planes simultaneously. If a block slides on two nonparallel faces simultaneously (Fig. as shown in Fig.I7. Let n. the direction of sliding is parallel to their line of intersection. be vectors normal to each of the . Denote the normal to the sliding plane by fi. \' '.t8. hXn. Therefore. (2.In this case the initial sliding direction is parallel to the direction of the orthographic projection of the resultant force (r) on the sliding plane P. 2 Single-face sliding.s3) where the symbol lll means "is in the same direction as. 2.18 Sliding direction under single-face sliding.

and 0 if/is zero. Let sien (/) signify *1 if/is positive.e. The analyses of block stability under the action of self-weight only will be examined as a special case. x or) (2.Computation of the Sliding Direction Figure 2. Then the direction (s) of sliding along the intersection of planes i and 7 is s lll sign [(n. the resultant force on the block is . . 2.s4) n2 Figure 2. The sliding direction under gravity alone. The actual sliding direction is the one that makes the smaller angle (i. s /l/ nt Xn" ble-face sliding.I iflis negative. and Pr.20 Sliding direction for dou. X nz). less than 90") with r (Fie. r](n. sliding planes P. x nr) . Without other forces.20).. The direction of sliding (s) is either the same as (nl x nz) or its opposite (-n.19 Sliding on two planes.

s6) For sliding on face P.3106. The equation of the plane is AX+BY+CZ:D From (2. Since (1. giving D : 1. 2 r : (0.A. ^B : 0. .3106 .48 Description of Block Geometry and Stability Using Vector Methods Chap.32139x + 0..l | s lll w(Arc.2. ft. x is east. y is north. Cr) in (2. The Equation of a Plane Consider plane P with dip angle d : 30" and dip direction F :320' clockwise from north. and z is up.5$: *t2 fi.1). For a block fall.0) 0 0 -wl * . A: -0. BtC. it satisfies its equation..86607.: -B. Ct AJ BI CJ glvlng S lll [sign (Afi.53).58) EXAMPTE GATCULATIONS ln the following examples. 2l and (ft. Bt.W 0 At Br C. Example 2.and C: 0. -W) (2.38302Y + 0.W A.55) where W is the weight of the block (W > O).2. -w) (2.32139.cb -U? + B?)) (2. alone.B)j(B'C.866072 : 1..55) together with fi..Bt . . : (Ar. .38302. Then the equation of plane P is -0. I W(-8. we substitute in Q.Afi. 0. we must substitute r from (2.AtC' .7). The plane passes through point (I.s7) For sliding simultaneously on planes and j. the sliding direction must then be s lll (0.1. 2l fi1 xt: At Bt C.xi)xfi. AlCy A. A.1 xfi1 : At B. L) is on this plane. B. so I + 28 * C : D.) (2. 0.

Xt : 2. is parallel to the line of intersection of Pr and Pr. and L. Equation (2. 0. Yr .10) determine the radius vector (X.3.C: l. 0. 0. From (2. -0. Pz. where Lr. -2) in direction (2. and Pa Assume that a tetrahedron is the region common to the intersections of 21.43301.23631) Example 2. The Intersection of a Plane and a Line Consider plane P. has &z: 60" and Fz: 150" From (2.12).75000.Example Calculations 49 Example 2. A:2. and fi.9).X0: l. Then equations (2. 1. A Tetrahedron Greated by Planes Pr.-2.7) the unit normal vectors ft. Ls.7 6494. Y. and Zt: 3. L".059391.Yo:L. has ilt:20" and Ft:280o P.2.33682. D:4. from the origin to the point of intersection. whose equation is 2x+3Y+z:4 and a straight line passing through (1. I rz : (0. by X:L!2:3 Y:t-2:-I and Z: -2 + 3 : 1 Example 2. are the half-spaces below planes 1. -2. 0.Zo: -2. Z).3). L2. ate fir : (-0.50000) vector ltz: ftrx h. The Intersection Vector of Two Planes Assume that P. givinEto : 1.2.3) gives X:l*2t Y:L-2t Z:-2+3t The piercing point of the line on the plane is found by making the following entries in (2. and 3 and Un is the half-space above plane 4.0. P3. and (J 4. These planes are defined by values of d and f as follows: Plane o(deg) f(deg) 14590 2 4s 330 3 45 2lo 4090 .59918.8:3.4.93969) and ftz : (0.

61237Y + 0. the block has exactly four corners. C1r.61237Y + 0.35355X .Crzt : (1.7320 V:*l"ll tl I -r.707t02: 0. x: Crzr . The Area of Every Face Consider further the tetrahedron of Example 2.707102 < 0. These ate Cp3.zi -0. Similarly.35355X + 0. The area A of the face in plane Pt is the area of the triangle with corners Crro all sharing the common index 1. 0. Let r: CtzzCrrn andb: CpsClsa [see Fig.2.ll I r.2.Zl).70710 The solution is Ctzs : (0.707102 :0. t.0.35355X + 0. and3. and 4.353ssx .21(a)]. The volume of the block can now be computed by entering the coordinate of Ct zzt Crz+t Crrn.732.732.23).70710X + 0r + OJ07t02 < 0. with only four planes.4. 1).61237Y + 0.70710 (a) L.50 Description of Block Geometry and Stability Using Vector Methods Chap. and (d) is not necessary.n. 0) In this case. is the intersection point of planes I. L. and C.5. -L) . 1.6L237y + 0. so a selection test using (a).70710 _0. According to the discussion following Q.707102 < 0. 2 The half-spaces are then Lt. .0.7320 0lt:1. giving lttl 0 0 1l .Cr"o. 0) and Czst : (-2.70710 and -0. 0. and Czz+ in Q. -1.0) Using the method for the other corners gives Ctb* : (1.70710 (c) U+i 0x+ 0r+ Z>O (d) Now we must compute the coordinates of every corner C. 0. It is found by the simultaneous solution of 0.7320.70710x + 0r + oJo7Ioz: 0. -I) b: Cr s4 . (c). Simultaneous solution of the corresponding three equations gives Ctzt : (1.7320 0l ll -2 0 0l Example 2.1r of the block. (b). . 2. Clrn is the intersection point of planes 1.70710 (b) Lsi -0.7320.Crrr: (1.

l) .2.6. fir : (Xt.s a.73201 ) :2. Yr. of the face in plane 3 is the trangular area between Cr Lgt Ctgqt and C2. 1.7) ftz : (Xr. Ar: Ila x bl : +ll r'7320 -t l' * -1 1r -r L . and Crrn.rrCrrn X C.1961 Example 2.23). and Crrn. the area A.rnCrr+l:5. The Angle between Two Vectors Given two vectors fi. As: tlC. {b} By (2.4494 and finally. The area I. of the face in plane 2 is the triangular area between Cr"r. 2. Az : + lC.7320l'zltrz [| -r. .rrCrr+l:2.rCrrn X C.4494 Similarly.zrzo -11 -1 1l 1 -1.t and fi2.21(b)].4494 [see Fig.8. . Zt) : (9. A+ is the triangular area between C1 .21 Area of a triangular face.C134. Zr) : (1. Crrn. 51 Face 1 Crzs a Crzc 2 (a) c'"o 1 Figure 2. Yr. A+: *lCrrnCtrn X C.rCrr+l : 2.n.

0.38302.26): cos d : 0. 0.13302.4131 7.t and frz is calculated from (2. we compute the upward unit vectors of these planes: fir : (0.7 6604) fit : (-0.9. Fr:160o. l).38302. Example 2.It is then 57. 2. Using (2. 0.984802 : 0 .0.8.86602) The angle between fi and y is calculated using (2.7.41317 X + 0. Example 2.&278) Then use (2. 0. 0. This is the complement of the desired angle. First compute the normal vectors fi.349". giving cos d : O. The Angle between a Plane P anda Vector v Given planeP with d : 30o. respectively. -0.64278) ft+ : (0.0.651".17101X + 0. and fi.I3302X .7).32139.75440. 0.766042 : 0 Psi -0. 0. and Pn in the block pyramid are Pri 0. F : 320".93793.0.754r'. The Angle between Two Planes Given Pr with dr : 30". and PL with d2 : 50o.7) to compute the normal fi of plane P. 0. Pr. Example 2.030153Y + 0. d+:10 f+: 80 We will compute the block pyramid created by the intersection of half-spaces u r. :320".653.7t984. Find the Block Pyramid with Four Planes Given four planes: Dip Angle Dip Direction Angle (dee) (dee) Pt dt:30 Fr: 90 Pz dz :4O Fz :320 Ps dr :50 fs : l9O Pq. i : (-0. First use (2.26). : e032139. of P.26200. giving 6 : 20. vector v : (1. -0. Lr.9675 and d : 78. 0.17101. 2 the angle (d) between fr.49240Y + 0.84480 and d : 32.49240.030153.642182 : O P+: 0. Lr.292".52 Description of Block Geometry and Stability Using Vector Methods Chap.86602) fiz : (-0. which is between v and plane P.0Y + 0.98480) The equations of planes Pr. Pr.866O22 : 0 Pz: -0. 0. and Pr.5X + 0f + 0. f .26): cos 6 :0.5. and u n.86602) fir fiz : (0.

4. The intersection vectors from :[Ir/ that satisfy simultaneously all the inequalities (a).Example Calculations 53 The terms right of the equal signs are zero because the planes Pr. *Irs. The equations of half-spaces Ur . Y. i+i Irz : e0.015076) Izs : (0.43659. Lr. -0.984802 > 0 (d) Compute all of the intersection vectors: ItJ : fi4 x fi1.89442.53790. and (d) simultaneously.0. 0. 0. Pr.642782 <0 Ls: (c) Ui 0. -0..53790. (b). Referring to Example 2.lrn.642182 <0 (c) -0.1O. *Ir+.13302X . and (/n has only three edges. Lr.s0000x+ (a) Lz: -0.0.17101X + 0.e.00000r+0.096664) Ir + : e0.866022>0 (a) Lz: -O. one without any edges in space) means that the intersection of these half-spaces will always create a finite block.126113. we test all of the vectors -flrz. -0.37720) Ir c : e0. Lr.49240y +0.89442.0.2. 0.9.75440Y + 0. and U n are Lr: 0.096664) -frn : (0.46182. and Pn. Izr: (0.37720) lz+ : (0. tlrr. -0. which have the same dip angles and dip direction as in Example 2. irj : 1.13302X - n: (J 0. Lr. (c). 0. 0). and flrn one by one. and Pn pass through the origin (0. -0. the equations of half-spaces Lr. and (d) are Irr. we compute the block pyramid created by Lr.74084. Lr. (c).5X + 0r+ 0. Example 2.37720) Izt: (0.24fr92.766042<0 (b) -0.7 6232.24620) Irr : (0. Lr. The method of this example will be used frequently to judge the finiteness of rock blocks. P2. 0. : (X. -0. Pr.34430.17101X+0.030153r+0..766042 < O 0A924frY (b) Lti 0.3. Lr. Zj in equations (a).16368.9. 0. 0.866022<0 0. So the joint pyramid cut by (Jr. Lr. 12500) Finally.65333.16368.030153 r + 0. (b). and Un. -0.75440Y + 0. Given four planes Pr. Determine That a Block Pyramid ls "Empty" In the next chapter we will learn that an empty block pyramid (i.413I7 X + + 0.413r7X+ 0.76232.12500) These three edge vectors completely define the corresponding joint pyramid.0.984802>0 (d) .24092. and U n are Urt 0.46182. t[z+.42643. and -Irn. 0. P". -0. substituting the coordinates of fI.

11.Lr. (c). satisfies simultaneously equations (a). the blockpyramid cut by Lr. l) Then the resultant r of forces w. arising from weightt : w (0.0) and from inertia i € : ( Then the sliding direction is Ji . l) rhe sriding nt*il i. Gompute the Sliding Direction for Uplift Here we assume it is known that the resultant force tends to cause uplift. dip directi on p : ze. Suppose that the resultant is r : (0. Example 2.:f:Til ' r : (o' 0'6' o'8) from formula (z. Finally. None of the vectors fI. Therefore. -flr+. 0.tJl . *lzo. and +I34 one by one. (b)..2. Gompute the Sliding Direction for the Gase of Sliding on a Single Face Suppose that the resultant is t : (I.26200. -5) from water: p : (4. and (d).^o The unit vector of plane P is fi : (-0 . 3. and (d). and e is J:lr*p*e giving r: (6. substituting the coor- dinates of f Ir7 : (X.7 tgg4.54 Description of Block Geometry and Stability Using Vector Methods Chap. (c). Z) into equations (a). p. 0. *lzt.12.13.4. -4) Example 2. *Irr. i+j (this is exactly the same as in Example 2.50o.64279) . Example 2. Gompute the Resultant of Forces Suppose that there are three forces acting on a rock block. we test all of the vectors *lrr. Y.1 xfiy i. 2 Compute all of the intersection vectors: lrl:fi.0.3.j:Ir2.2.9).Lr. (b). 3. such a pyramid is said to be empty.and Un has no edge. 1.

arc fir : (-0.29593) Example 2.50000) From equation (2.78181.. then i: (0.73446.7017) @ x f) x fi: (1.3626. _1. -0.73446.33682.57531. The orientations of these faces are given by Pra a"r : 20". x fir) . and Pr.54880.54) we know that the sliding direction unit vector ^f is s lll sign [(fi' x ft.53) we know that the sliding direction unit vector J is .0.22690) (fi.93969) fiz : (0.flll (. r](ft.76494.Example Calculations From equation (2. x fir).59918.14. -0.0.then ^i : (-0. 0.23631) . -1. -0. -0. -0.1) The unit vectors of planes Pt and P.f is the unit vector of (fi x l) x fi. x fir): (-0. rl : -1 sign [(fi.r)-rl(fi' x ftr) i is computed by the following steps: fit xfiz: (0. r: -0.0235. fz: 150" The resultant force is given as r: (0. x fir).axi)xfi fr x i: (-0. rl(fi. 0.8829.0.34841 sign [(fit x fir). Gompute the Sliding Direction in a Case Where Sliding ls Simultaneously on Two Faces Suppose that the sliding faces are planes P.71'270) .75000.059391. ft :280" Pz: &z: 60".43301. 0.22690) i is the unit vector of sign (fit x fi"). 0.

compass. The tools required for the creation of a plot. they can be adopted in part as a complement to vector analysis to provide a semigraphical solution technique placing the entire computational process inside a glass chamber for observation. pro- tractor. Our approach is to make use of the power of a modest cal- culator-the common. besides the calculator. some of which are listed with the references. and scale. Stereographic projection as a graphical device for solving geological. However. these plots can be drawn with computer gra- phics on a video screen or plotter. while some of the essential material already available in this literature will be repeated here. most of the content of this chapter is novel. and other spatial problems has been discussed in a number of books and articles. The techniques discussed are complete in themselves and can serve as a full substitute for the previous theory. The thread that connects the sections of this chapter is that stereographic projection procedures can be refined and enhanced by supporting them with initial calculations. chapter 3 Graphical Methods: Stereographic Projection The use of vector analysis to describe and examine blocks permits speedy solution of real problems by means of computers. faithful companion of almost anybody who reads this book-to establish the points and lines to be plotted. crystallographic. are paper. it will 56 . The use of a stereonet is therefore not necessary. Alternatively. An alternative solution method incorporating the stereographic projection is presented in this chapter. Before plunging into this procedure. although it may be preferred by custom or previous bias. On the other hand. The use of graphics to examine the geometric relations in projec- tion at any stage of the computation offers a clearer perception of the geometric and physical relations being manipulated inside a computer. pencil.

and position by three orthogonal projection planes (views). (Warren J. ofwhichstereographic projection is an example. An example of orthographic projection of a three- dimensional object is given in Fig. Reprinted by permission Prentice-Hall. Projection of Distances Orthographic. Inc.9 SURFACE "8" TOP 7 8. shape.1. nonparallel views. parallel and perspective. a polyhedral block is determined in size.1 Orthographic projection of a three-dimensional object using mul- tiple views.. renewed 1980. 3.1. 3rd Ed. @ 1952. Englewood Cliffs. FUNDAMENTALS OF ENGINEERING DRAWING: For Technical Students and Professional Draftsmen. The well- known orthographic projection belongs to the first type. Luzadder. N. This technique is probably the most usual basis for the drawings prepared by engineers. including orthographic and oblique projections. 3.9 8. TYPES OF PROJECTIONS Projections can be organized into two groups.J. in which the construc- tion lines transferring the points of the subject to the surface of projection are all parallel rays. Projections Types of Projections Types 57 be instructive to place sterographic projection into context with other graphical techniques.7 I FHONT RIGHT SIDE POINT ANALYSIS SURFACE "C" SURFACE "TY' Figure 3. gathers the construction lines together at one or more foci behind the surface of projection. SURFACE "A" 1.) . The lines and planes defining an object are transferred to the drawing by means of rays drawn perpendicular to the projection plane. In Fig.. These lines and planes are fixed uniquely by multiple. The second class of projective techniques.

PENCILS SHOW THE DIRECTION OF PROJECTORS . 1952. This method proves useful in conjunction with orthographic projection to provide inclined views of objects.) Perspective. Figure 3. Imagine all directions (unit vectors) radiating from a central point. 3 Oblique. Graphical Methods: Stereographic Projection Chap. Perspective projection of this type is quite useful to convey shapes of objects but is not appropriate when it is intended to convey measurements of distances or angles. Figure 3.2 Orthographic and oblique projections. (From Luzadder. with permission.hfi Qe Figure 3. Otthographic Projection of Angular Relations The orthographic projection can be used to show the relations between lines and planes in space and to measure the angles between them. This bundle of unit ..3 shows such a projection of a complex polyhedral block.2 shows a related parallel projection technique in which the construction rays are oblique to the plane of projection. Nonparallel construction rays are used to create three- dimensional views hpving the appearance of true shape.

of Projections 59 Types PIERCING POINTS PICTURE PLANE GROUND LINE Figure 3.4(b). Small circles of the reference sphere that are generated from a family of cones about the axis of the great circles will project to the diametral plane as straight lines (lines of latitude) as shown in Fig. Furthermore. I in Fig. These project to the diametral projection plane as a family of elliptical curves.4) is a specific point on the surface of the sphere. Orthographic projection produces u piun of a diametral plane of the sphere by means of construction rays directed perpendicularly to the diametral plane.. It has the disadvantage. Thus unit vector OA is projected to point Ao. A series of planes inclined with a common line of intersection through the center of the sphere creates a family of great circles on the reference sphere. because of the crowding of the lines of longitude near the edges of the projection. An even stronger reason to reject orthographic projec- tion of the sphere for the development of block-theory graphics is its failure to . (From Luzadder. however. They are analogous to the lines of longitude of the globe.g.3 Perspective projection. as shown in Fig. 3. 3. measurements of interplane angles can be inaccurate. 1952. that equal solid angles can produce greatly unequal areas on the projection.4(b). Orthographic projection of the sphere is used commonly for cartography. with permission') vectors produces a sphere and the tip of any one of them (e. 3.

60 Graphical Methods: Stereographic projection Chap. 3

ffia/ |
v//// \ \\\\\
\ \ \\\\
// I I \ \ \\\\
4 '// / I \ \\\ \
tl \

Rsina As


t\ II
\t I
OAo=Rsina \ \\\ \ / lt
\\\\ \ \ I I / ///Z
(a) \\\\ \ \ | / / ,/.//z

Figure 3.4 Orthographic projection of a reference sphere: (a) basis for the
projection; (b) a projection net of lines of longitude and latitude.

distinguish symmetrical points in the upper and lower hemispheres. If line OA
is inclined o with either the upward vertical or the downward vertical, distance
OAo from the center of the projection to its representation, point ,,{0, will have
the same value, R sin d, where R is the radius of the reference sphere. A
perspective projection technique will be required to differentiate symmetrical
points in the upper and lower hemispheres.

Equal-Area Projection

Figure 3.5(a) shows a homogeneous projection of the sphere called equal-
area ptojection. Line OA, piercing the reference sphere at point l, is projected
first to a tangent plane at the top of the sphere along a circular arc whose center
is at O' at the top of the sphere. This projection on the tangent plane of the
sphere is then produced on a parallel equatorial plane by construction rays to
a perspective point (F) at the bottom of the sphere; the representation of line
OA on the equatorial plane of projection is then point Ao. If line OA is inclined
at a with the vertical, distance OAo will be equal to RJT sinul}.
The advantage of equal-area projection is its homogeneity, meaning that a
solid angle produces a unique projection area (but not a unique shape); that
is, the area of its projection is the same anywhere on the sphere. This property
facilitates statistical operations with lines. However, great circles and small
circles of the sphere [Fig. 3.5(b)] project as nonconic loci that are difficult to
draw. In contrast, the stereographic projection produces loci that are the ulti-
mate in simplicity.

Types of Projections 61

OrB = OtA, = ZrrD.A ti"t
oAo = .'fr.a"i"$


Figure 3.5 Equal-area projection of a reference sphere: (a) basis for the
tion; (b) a projection net of lines of longitude and latitude.

StereograPhic Proiection

Figure 3.6(a) presents the basis for stereographic projection, which we
will examine closely in this chapter. Consider line OA inclined at angle a with
the vertical. The piercing point A of this line on the sphere is projected to a
horizontal equatorial plane along perspective line AF, with the perspective
point (r) located at the bottom of the sphere. The point ,,{o representing the
projection of line oA is located at distance R tan ulL ftom the
center of the projection Plane.
Figure j.O(U) shows the families of great circles and small circles corre-
sponding to lines of longitude and latitude, as discussed in the previous figures.
All these lines are portions of true circles in the projection plane. The proposi-
tion that any circte on the reference sphere projects to a circle in the projection
plane is proved in the appendix to this chapter. This important property makes
,trr.og.uphic projection far easier to draw and formulate than either ortho-
graphic or equal-area projections.
Note from Fig. 3.6(b) that a given solid angle will project with different
areas in different regions of the sphere; that is, the projection is not homogene-
ous. In this ,.rp..t, its quality is intermediate between orthographic and
equal-area projections.
Compared with orthographic projection, both stereographic and equal-
area projection methods will produce a unique point corresponding to every




Figure 3.6 Stereographic projection of a reference sphere: (a) basis for the projection;
(b) a projection net of Iines of longitude and latitude.

Types of Projections

unique direction radiating from the center of the sphere. Thus there is no con-
fusion between symmetrical points in the upper and lower hemispheres.
Figure 3.7 contrasts a stereographic projection embracing a focal point
at the top of the reference sphere with a stereographic projection having the
focus at the bottom of the sphere. Regardless of which focal point is chosen,
the equatorial plane projects as a circle, which is denoted the reference circle.
All points on this circle represent horizontal lines. If the focal point is at the
bottom of the reference sphere, as shown in Fig. 3.7(a), a line Iike OA that
pierces the sphere in its upper half will project to a point that is inside the
ieference circle [Fig. 3.7(b)]. If the top of the sphere is taken as the focal point,
a point like OB of fig. 3.7(c) that pierces the sphere in its lower half will lie
wittrin the reference circle, as shown in Fig. 3.7(d). Because of these properties
of the stereographic projection, the upper and lower foci are said to generate,







Figure 3.2 Upper- and lower-focal-point stereographic projections of a line
und uplane: (a) section of thereference sphere; (b) projection plane; (c) section
of the reference sphere; (d) projection plane. (a) and (b) give lower-focal-point
projections of a line; (c) and (d) give upper-focal-point projections of a line'

64 Graphical Methods: Stereographic Projection Chap. 3

respectively, lower and upper hemisphere projections. This can be misleading,
however, as either choice of focus permits projection of lines in both hemi-
spheres. In the case of the lower focal point, lines contained in the upper hemi-
sphere will project inside the reference circle and lines in the lower hemisphere
will fall outside the reference circle. tn this work we adopt the lower-focabpoint
projection, tmless otherwise indicat ed.


We have seen in Fig. 3.7 that stereographic projection of a sphere permits the
construction of points and loci representing lines and planes in space. Remember
that all such lines and planes are assumed to pass through the center of the
reference sphere. A line will then project as a point somewhere in the projection
plane (except for a line from the center of the sphere to the focus of the projec-
tion)' A plane passing through the center of the reference sphere generates a
great circle on the surface of the reference sphere and projects to the plane of
projection as a circle. Several procedures for constructing these projections are
described in the following sections.

Stereographic Projection of a Vector

Consider the vector v in Fig. 3.8(a), with coordinates

v: (X, Y,Z) (3.1)
Orthographic projection of v on a horizontal plane (xy) gives vector u' in direction
f with y; v makes an angle a with the vertical (z). The stereographic projection
of v is constructed by means of a vertical section through the reference sphere
along the direction of r.r'. This section is shown in Fig. 3.8(b). We arbitiarily
assign the reference sphere a radius R. Then the distance OV to the projection
point of v is given by
OV: Rtan$ (3.2)

where, as noted, a is the angle between v and f. Point Zrepresentingvector v
is plotted in the projection plane [Fig. 3.8(c)] by its polar coordinates (o V, p),
where B is measured clockwise from north (j).
Alternatively, V can be plotted from its Cartesian coordinates (X", Y").
In a vertical section of the reference sphere along the x axis [Fig. 3.8(d)] we have
shown OV",the orthographic projection of v in this plane. From similar triangles


Similarly, a section of the reference sphere alongyz will give
R+ z R
T Yo

Stereographic Projection of Lines and Planes



y north

(c )


OV": orthographic

z projection of v
in the xz Plane
o xo




ligure 3.8 Stereographic projection of a vector.

Graphical Methods: Stereographic Projection Chap. 3

Given a point (x, Y, z) on the surface of the reference sphere, R2 : xz + yz
I 22, we can determine the stereographic projection of the point from its
Cartesian coordinates Xo, Yo in the projection plane, with

xo: R+ z
and (s:Fv
y- Rr

Let X : xlR, f: YlR, and Z: z1n rhen the coordinates xo, yo of the
stereographic projection of a unit vector (F, t,Z) are given by
rl RI
- l+z
z16: (3.aa)

and vRf
- (3.4b)
l al
- r+z

If the vector v establishes a point (X, Y, Z) not on the surface of the sphere
(i.e., R2 + xz + Y2 + z'), the length of v is proportioned so that it does. Then
the representation of v in the stereographic projection will be given by
' - JXr+Y2+22+Z (3.5a)

" -gJXr+Y2+22+Z

As an alternative to (3.5), when vector v is given in coordiate form (3.1),
we can plot its stereographic projection in polar coordinates (oV, a) with


and a: tan-t
# (3.6b)

Stereographic Projection of the Opposite of a Vector

A vector v : (X, Y, Z) has an opposite
-v : (- X, -y, -Z). In Fig.
3.9, the stereographic projection of v is point V, and the stereographic projec-
tion of -v is point V'. From Fig. 3.9(a),

OV: Rtanl (3.2)

and ov':Rtan(no-#)
Therefore, (ov)(ov'): pz (3.7)
Given any vector v, through the center of the reference sphere, we can calculate
OV from (3.2) and then determine OV' from (3.7).

making triangles p'CF and pCF isosceles. and the stereographic projection of OP' is p'.e) cos d . and OP'is its dip vector. line pp' in Fig.8) and |.10(a) is a vertical section through the reference sphere along the direction of dip. ate accordingly determined by OC: Rtana (3. R (3.10(b)]. 3. point C is the center. the distance p'C: distance Cp: distance CF because triangle p'Fp is a right triangle and line FC bisects its hypotenuse. To construct this circle. we must find its center and radius. Figure 3. Stereographic Proiection of a Plane A plane that passes through the center of the reference sphere projects as a true circle in the stereographic projection. that is. + Also. and whose radius.10(a) is an edge of the inclined plane. The stereographic projec- tion of OP is point p.9 Stereographic projection of the opposite to a vector. 3.10(a) offer simple formulas for the center and radius of the circle. r.3.POp:{P'Op':q Then < OFP: 45o . along the direc- tion OC in the projection plane [Fig. that { P'FC also equals 45" . It follows. Therefore. at C. Consider an inclined plane dipping a below horizontal in direction p measured horizontally from the x axis.10(b) is a diameter of the required circle and its bisector. First note that 4.Stereographic Projection of Lines and Planes 67 ) o x / V' {b) Figure 3. Line POP'in Fig.-. The angular relationships in Fig. then.dlz. so that < CFO: c. The stereographic projection of the inclined plane is a circle whose center. 3.

the coordinates will be (-C*. f . 3 OF=r o =45" -t CF=p'C=Cp=r (a) Reference circle Stereographic projection of plane IN il> l+ lx 5 |+ 5 {b) (c) Figure 3.cosp (3. normal to the .2) be an upward-directed unit vector.10 Stereographic projection of a plane.68 Graphical Methods: Stereographic Projection Chap. Let fi: (N.10b) These apply to a lower-focus (upper-hemisphere) stereographic projection of radius R. -Cr).If the upper focus is adopted. The coordinates of the center of the circle representing a plane with dip a and dip direction p arc C*:Rtanasinp (3.10a) C":Rtand.

JV.: R! 'z (3.S) or (3.13d) Figure 3.13d) and calculate r from (3. 3.1lc) Since Z>O and X. 3. 3.13b) and plot the center Cin direction p from y [Fig.9).13a) and draw the circle. then draw the circle from C with radius r [Fig. Calculate r using (3. Then find the strike line's projection points A and. (3.1lb) and Z: CoSd.\/Xzt ---:L-Yz (3. 1.f : sin usin B (3.12d) Substituting (3.11a) f: sin ucos p (3.B [Fig.12b) Then (3.13c) -Z c. given its dip (a) and dip direction (f) and reference circle radius R.10) gives the following coordinates for the center (C)and radius (r) of the stereographic projection of the plane: R (l.12) into (3.11 summarizes the various ways to construct the stereographic projection of a plane. +V _A -----. a 2z : l. Calculate OC with (3. 3. Calculate the coordinates of center C with (3.13a). Then .13a).9) or (3.rZa) .-.9) or (3. * f.10(c) as .13c) and (3.11b) give cos 'd : .11a) and (3. Calculate fi with (3. Find center c with AC : BC : R. (3.11(b)].-----=_-: -Z.11(c).d-..9) or (3.11c) determines the trigonometric relations according to Fig.tan.11(a)] as either end of the diameter of the reference circle in directi on p + 90".JFTTfr - (3. Z sin a . (3.Stereographic Projection of Lines and Planes 69 inclined plane that dips a in direction B.12c) and sinB:---L: . 3. 2.8). JX'+Y" (3.10) or (3. (3. . (3. and (3. Finally' draw a circle through C with radius r.l3a) a ---: Z OC: RJr-.13b) /1 _RX (3.

.12(a).f' represent the two directions along the line of intersection of P1 and Pr. whose intersection points I and. An example is shown in Fig. 3. 3 AC=BC=r=R/cose (cl Figure 3. 3. The stereographic projection of these planes [Fig.11 Alternative methods to construct a great circle. any two planes will have a common line. The Line of Intersection of Two Planes Since all planes considered here pass through the center of the reference sphere.12(b)] yields two circles.) has dip angle (a) equal to 60' and dip direction (p) equal to 100" and plane 2 (Pr) has a :50o and p: 260o. Graphical Methods: Stereographic Projection Chap. where plane 1 (P.Iis inside the reference circle (the dotted circle) and is therefore directed .

We will construct the stereographic projection of a small circle representing the locus of lines making an angle f with n [Fig. 3.13(b) shows a section of the reference sphere along a vertical plane through fi. it is denoted a small circle because it can be generated also by the intersection of the reference sphere with a plane that does not contain the origin. with stereographic projection points A and B.Stereographic Projection of Lines and Planes (b) Figure 3. This cone pierces a sphere about the origin along a circle. Figure 3..lyrng outside the reference circle. Ob. and is opposite to L A Small Circle The locus of lines (through a common origin) making an equal angle with a given direction through the origin is a cone. Then the radius of the small circle is . Consider a unit vector fi from the origin with coordinates fi : (X. the bisector of . The upper and lower limits of the conical locus are lines Oa and. into the upper hemisphere.13(a)]. is its center. is directed into the lower hemisphere.12 Line of intersection of two planes.{. AB is the diameter of the small circle in stereographic projection and C.8. Y. Z). I'. By the fundamental property of stereographic projection (Section 2 of the appendix to this chapter). any small circle on the reference sphere becomes a circle in the projection plane.

14b) Then r: t(oB _ oA) (3.14) in (3. Graphical Methods: Stereographic projection Chap.17) and simplifyrng gives oc: cos 4sina I -F cos c (3.15) and simplifying gives . 3 (b) Figure 3.Q and <OFB:ry OA:Rtan+ (3.laa) and OB:Rtanry (3.17) Substituting (3.13 Stereographic projection of a cone.18) ._ Rsind (3.15) Substituting (3.16) cos@ cos d The distance from the origin to the center of the projection is oC: *@a + OA) (3.14) in (3. r:AC:BC Since <OFA:n.

we can project the locus of lines equidistant from a normal fi to plane P as follows.----. (3. 3.:(X. calculate OC from (3.19a) and (3. (3.22) and the radius (3. substituting these in equations (3. the coordinate formulas for the center (3.20) - cosS *Z t-'=-------'----'- OC: 4X2+Y' (3.18) and lay it off along a line making angle a with x [Fig. Then use (3.22a) r_ uY Y (3.21: t Rsind ___________J_= (3.I2). 3. Knowing the dip (a) and dip direction (B) of plane P. and (3.11).13) all apply here. (a) {b} Figure 3.Stereographic Projection of Lines and Planes The coordinates of the center of the small circle are r _ Rsinasin/ vx-cosf*cosa (3.1a(a)]. 3. . and (3.20) are preferable [Fig.19) gives the following simplecoordinateformulas for the radius r and center C of the small circle of lines making an angle t' about fi..18). If you are using a computer.f.1eb) Since fi is the unit normal to plane P with dip o and dip direction B [Fig.13(a)]. (3.14 Alternative methods to construct a small circle.-.-----.16).16) to calculate r and draw the circle with C as center.22b) . (3.2r) cosS I Z c*: cnsSrl1/* Z -. cos6 + Z ln summary.14(b)].

Procedures for doing this are described by Phillips (1971).15 A stereonet. (b) for the entire sphere except a region about the position of the focus. We will discuss the construction of a stereonet as an example of methods (b) Figure 3. showing the lines of longitude and latitude of a refer- ence sphere: (a)for one hemisphere only. 3 Example: Gonstructing a Stereonet A stereonet is a projection of the longitude lines of one-half of the reference sphere. The net is used to obtain approximate readings of angles between lines and planes by tracing great circles and rotating the tracing about the center of the projection. and Goodman (1980). Graphical Methods: Stereographic Projection Chap. Hoek and Bray (1977). .

15(a) is a stereographic projection net constructed with d: 10o. and the small circles. is on one side or the other of P. Assume that P. together with p : 90o.90o.10) yields R '- y-_ coskd (3.24) . -R. representing the projection of a family of cones about the line of intersection of great circles. STEREOGRAPHIC PBOJEGTIOil OF A JOINT PYRAMID Stereographic Proiection of a Half'Space In Chapter 2 we defined a joint pyramid as the set of points common to all the half-spaces bounded by the plane of each face of a rock block when these planes are shifted to pass through a common origin. 0. All the joint planes plotted in stereographic projection satisfy the requirement that they contain the center of the reference sphere. In the stereonet.270' into (3. substituting these values into (3. fi.22) gives : Rtan kd r cr: o (3.0). so that Q: d. R. "'. . sin d cos p. Substituting these values. Figure 3. one finds two types of circles: the great circles. is always upward or . the stereographic projection can serve to represent joint pyramids. d. 0.. with the same value of R as in the hemisphere net.23) C* : *Rtan kd and C":0 The small circles arc a series of cones about fi : (0. : (sin w sin p. 0) or fi : (0. Let the angle between each successive great circle be equal to d.. using the formulas above to locate only those points within the reference circle.1 uY: +R cdkd Figure 3.Zd.. is a joint plane that passes through point (0. Any vector from (0. . it lies in either of two half-spaces created by P. Therefore. that is. representing the projection of a family of planes with a common intersection. Therefore. is fi.15(b) is a more complete coverage of the sphere.20) and (3. 0). Then the planes have dip 6 : akd with kd : 0. 0) that is not contained in plane P.9) and (3. the unit normal vector to plane P. cos c) (3'25) We will adhere to a rule that cos d > 0.Stereographic Projection of a Joint Pyramid explored in the preceding two sections. The angular increment is kd. . If h dips a (0 < d < 90") in direction B (0 < B < 360'). 90o.2d.

has no part in the bounding of the regions. [n the lower focus projection that we have adopted. x : 0 [as in (2. The corner points of a region are then the .0. 0). Each numbered region in the stereographic projection can be thought of as a set of radius vectors inside a particular joint pyramid.16 Lower-focal-point stereo- graphic projection of half-spaces of plane P1. as the set of all vectors x obeying fi. is similarly.16). The stereographic projection of a plane. Similarly. 0.32)1. i: L to n. The system of planes P.. Each of these planes is represented by a great circle on the reference sphere and therefore by a great circle in the stereographic projection. 0) to points above P.0.0). cuts the whole sphere into a number of pyramids all having their apex at (0. the terms "upper" and "lower" may seem arbitrary. Figure 3. On the figure. (Fig. 3.26) The lower half-space (L. is determined by one of the two horizontal lines normal to P. its equation is ff..x)0 (3.27) When P. is a vertical plane. We will define the upper half-space (U. Pr. each determined in orientation by a plane. the half-space U. 0. The reference circle. but is shown for clarity. the set of all vector x obeying fir. contains point (0.. 3.26) and (3. . is a great circle. the half-space below plane P. generates a series of regions on the projection plane. Since P. as shown in Fig. But they are still adequately defined by (3.17a. The intersections of all these circles... 3 horizontal. The Intersection of Half-Spaces to Form Joint Pyramids Suppose that there are n nonparallel sets of joints.) of P. the dashed circle on the figure. is all the region outside the great circle of plane Pr.) of P.0). these regions are numbered arbitrarily. P.27) when fi. passing through the origin (0. is the region inside the circle of plane P.x(0 (3.76 Graphical Methods: Stereographic Projection Chap. of all vectors from (0.

a'1"'. using a lower-focal-point projection.Stereographic Projection of a Joint Pyramid 77 t/ Lt4 / N P.. to \ t4 11 .' a. S7 (c) Figure 3. (d) the lower hemisphere. . using an upper-focal-point projection.N P.-.17 Stereographic projection of joint pyramids: (a) lower-focal-point pro!rction of the whole sphere." ---i---'\---j '\. (c) the upper hemisphere. 3Pt Ir')'. (b) projection of UrUzUtUc (region 1).P.

18 Normal to a plane.17(c) the region inside circle P. region 1 on Fig. In Fig.18(a).17(d) the region inside the circle of P. 3. fig. projected from the upper focal point of the reference sphere. 3. and Fig. fts corners are the projections of the four lines of intersection. <oFA =qs" -Z (a) Figure 3. 3.Inr. 3. Figure 3. I2a&rrd. An alter- native representation can be made by separating the upper and lower hemi- spheres in two separate figures. the plane is seen as diameter PP'. 3. thus we prefer to project with a single focal point. and these form the edges of the pyramid as shown in Fig. We have found it easiest to plot all the regions of the sphere from a single focal point at the bottom of the reference rphrrc. In Fig. That is. For example.17(a) is an upper-hemisphere stereographic projection of the joint pyramids. This can cause unnecessary confusion. ADDITIONAT CONSTRUCTIONS FOR STEREOGRAPHIC PROJ ECTION Several additional procedures will make it possible to carry out all of the steps eventually required in the graphical representation of block theory. Izs. a vertical section through the refer- ence sphere along the dip direction of P. Irs. 3. 3. . .17(c) is the upper-hemisphere portion. projected from the lower focal point.17(d) is the lower-hemisphere portion. belongs to half-space t/. Graphical Methods: Stereographic Projection Chap. we wish to find the projection of its normal fi. (In. The normal to a given plane.17(b).. and. Given plane p. (f z. (f s. 3 stereographic projections of the edges of the corresponding joint pyramid and the circular arc boundaries of a region are the projections of the faces of the pyramid.17(a) is the joint pyramid formed by the intersection of half-spaces (/. whereas in Fig. belongs to the half-spaceZ.

Z represents a unit vector 0 : (X.8).R2 dz'- (3.33) by (R' + XA + Y&)|QR). of the great circle. Suppose that we are given a point V: (Xo. Solving (3.28) gives : d: l.R (3. The normal is at point i[. distant d from O along OB. This diameter is AB. andthe radius of the great circle r : ABl2. Combining these two relations gives oc:#ryF (3.18(b).29) for r in terms of d yields :*_!4p .Inserting this in (3.33a) ^:I{2+XT+yX fr zRYo t : pzlj{yEo (3. The plane normal to a given line. oN:d:Rtand' 'T:n1l^ ll-cosd 1a. 3. at the center of the reference circle. Z) and therefore Xz+yz+Zz:l (3. First draw the diameter of the great circle that passes through O. Yo) in the stereo- graphic projection.ouo (3. and the distance to the center OC.2e) ON Rnl vr-l-R Suppose that we have a great circle in the projection and want to plot the normal [Fig.33c) I{2+XT+y4o Multiplying equations (3. Because < OFN : ul2. The procedure described above can be reversed to draw the projection of a plane given its normal.4) apply. -. 3. cos d : Rlr.33b) and + R2-X3-Yt z': (3.32).9). we obtain . these can be inverted to i zRXo (3. The vector represented by a point (Xolro) in the stereographic projection. . tan(ulT) : dlR.32) Also. r. so we know that d: ON. f . With (3.28) From (3. Suppose that we are given point N in Fig.18(b)].30) - From (3.31) Thus knowledge of d permits calculating the radius.28). n 2tan@12) OC:Rtatr'--n7-1nP616 and from (3.Additional Constructions for Stereographic Projection 79 inclined d with horizontal. equations (3.

37) gives (3.3aa) Y:Yo (3. .3a) permit us to cal- culate their corresponding vectors vl : (Xr. Yr. Zr) and v. 3. anywhere in the stereographic projection. Y. and Z.XrZr. Ir) and Vz: (X".38b) In summary. Assume that we are given two points V.2R (3.2. XrZ.13d) give the coordinates (C* and Cr) of the center of a great circle representing a plane with unit normal N. from (3. fhe ratio of two coordinates of a unit vector equals the ratio of the corresponding two coordi- nates of any vector parallel to and in the same direction as the unit vector. (3. we can rewrite (3.19. has coordinates given by ll : Yr X Yz : (YtZ" . Z) and its unit vector (t.38a) vy-Wn /1 -XrZt-XtZ.. Yr) and. V. . and V\.YrZr.V|is the opposite to point 2. Z) both have the projection point V:(Xs.13d) as follows: /1 RX W*: -T (3.35b) The normal ft to the plane common to y1 and v.37b) Substituting the cooldinates of fi from (3. An alternative graphical procedure to construct the great circle through two points is to find the opposite to one of them.35) and calculate the coordinates of the center from (3. 3 X: Xo (3.80 Graphical Methods: Stereographic Projection Chap. Zr) with 7 -R2-X?-Y'.13c) and (3. It is often necessary to find the great circle through two points in the stereographic projec- tion. non- parallel vectors.36) into (3. ?. .Yr). XrY.3ac) Vector v : (X.XrYr) (3.36) Equations (3. Yr.38). Therefore. : (Xr.35a) and 7 _R2-X?-Yi 22--____zF. Vr.34b) and Z. : (Xr. f .Rz-X'"-Y'. compute Z.13c) and (3. The center of a great circle through two points.37a) and Cr: V (3. (3. Yr) in the stereographic projection. Then the circle is constructed through I/1 . This corresponds to finding the plane containing two intersecting. Yr). pr----T (3. : (Xz. to find the center of a great circle through points Vr : (Xr. Equations (3. In Fig.

Additional Constructions for Stereographic Projection 81 Figure 3. we will find the trace of the line projected into the plane by parallel. . To find the orthographic I I I .19 Great circle through two points.t " 'tQ . This trace is the line of intersection of plane P with the plane that contains both v and the normal to P... orthographic projection. point Zrepresents line v and the circle for plane P is shown.\ LV \ \ I N\ l'\R"t fc eference circle )' _v' Figure 3. Given a plane P and a line v. 3.20 Stereographic projection - of the orthographic projection of a line ane on a plane.20. The orthographic projection of a vector on a plane. In Fig.

l[and Zusing (3. 3 projection of v on plane P: 1. Z is the projection of a vector v and P is the projection of a plane. This great circle intersects the circle for plane P at points e and e'. arc two planes. and.21(a). Plot the normal (N) to plane P. Construct the normal N to plane P.Zr)andAr: (Xr. Measurement of angle between two vectors. . and P. and Q' is its opposite. and Vr: Vt:(X0. P. 2. The orthographic projection of v on P is point Q.3e) The angle between two planes.33). In Fig.. The required angle is the complement of < NV.f r.21(b). (2) then using equations (3. and (3) calculate d from cosd:Irtr*7rfri2rZ. yor) and Vz: (X_or. and vr. Graphical Methods: Stereographic Projection Chap. There is another. 3. In of two vectors v. 3. Draw a great circle through V and N. calculate 0r: (Fr. simpler way to mea- (a) (b) Figure 3. .33) and (3. or betw€€n a line and a plane.l/.21 Angles between planes and lines.Yor). (3. The angle between them is the angle between their normals. Then determine the angle between . Given the projec- tion pointsvrand v.39).Zr). to find the angle (d) between them: (1) measure the coordinates of Z.

Single-face stiding.22 Sliding directions for single-face sliding. The angle between them can be read with a compass between the trangents to Pt and P. using vector analysis.i is then identical with the direction of l. In Fig. 3. (This property is established formally in the appendix to this chapter. and P". at either point of intersection.22(a) i is the direction of the resultant force and sliding is along plane P. If the resultant force r acts in a direction such that a block tends to slide along one of its faces. PROJECTION OF STIDING DIRECTION The direction of sliding of a block under a given set of forces was discussed. Suppose that the direction I of the resultant force applied to the block is oriented such as to lift the block from every joint plane. in Chapter 2. This direction can be determined by stereo- graphic projection procedures as well. . as shown. Construct N Figure 3. The stereographic projection has the property that the angle between two planes is exactly equal to the angle between two tangents to the great circle projections of the planes. constructed at their intersection.21(b) the two planes are represented by the great circles P. Lifting. 3.Projection of Sliding Direction sure this interplane angle. the sliding direction is parallei to the orthographic projection of r on the plane of that face.) In Fig. The sliding direction .

r) of circles P. -1).l/ normal to P. 3. that lies in the half-space of P. However. Thus s is f. as is the point f. In the frequent special case where the direction of the resultant force is that of gravity (0. and Pr. that is contained in the same half-space of P.2 and -. The sliding direction is the choice of Ir2 and -. Also.23(a). the normal to P lies along a diameter of the reference circle extended through C the center of the circle for plane P [Fig. 3 .. s. . But P. is the choice of st and s.I1. f is the projection of the resultant force direction. as shown in the figure. intersecting P at sr and sr.I. P2 (a) (b) Figure 3. Construct the plane P. The sliding direction. andPr. The plane common to i and Nis vertical and is therefore a straight line in direction OC. In Fig. we know that the half-space of the P. Sliding is on planes P. Then construct the plane common to N and f .r. is the intersection of the extension of OC and circle P.0. containing f. f cannot be plotted.22(b)]. f is downward and cannot be plotted. 3. whose intersection lines are the two crossing points (/. The sliding direction is then s. whose normal is at f . ^f.43 Sliding directions in double-face sliding. Graphical Methods: Stereographic Projection Chap. Sliding in two planes simultaneously. Now construct plane P. circle that contains i is the region outside the reference circle.. perpendicular to f The sliding direction. Under only its own weight.

1) and the radius R of the reference circle R:1 Using formulas (3.0. Stereographic Proiection of a Unit Vector Given a unit vector 6: (0. the lower hemisphere.38683.2.0.: 42o and dip direction f : L44". Therefore. Example 3. R:1 the coordinates of the projection point Z of the vector Y are y: (0. y: (0. Y:2.a) with .57975) Then we can plot the projection point V of v in the projection plane.51335.:53" Erample 3.f : 0. EXAMPTES Example 3. s is the line of intersection of planes 1 and 2 that plots outside the reference circle [Fig. that contains I is the region outside the reference circle.76604 We calculate the coordinates of the projection point V of unit vector 8. that is. f :0. Z:1.7 6ffi4) Let R:1 Using formulas (3. let the radius of the reference circle be unity: R:l . Z:0.1.23(b)]. Another method to plot point Z is to use formulas (3.Examples is the reference circle and the half-space of P.29999. by which OV :0. Stereographic Proiection of a Plane from its Dip and Dip Direction GivenaplanePwith dip angle il.8391 and d.51335.6).21903. 3. Stereographic Proiection of a Vector Given a vector y : (1.29068) so we can plot point Z in the projection plane. 0.5) with X:1.

and r I.3456 [Fig. 3.728{/.20710) and .61237 Y :0.11(c) Example 3.35355. 3. i: (0.11(b)l 3. Using C. r : t.R : I we compute the data of the projection circle of plane P as follows.35355 and Z : 0.90040 Alternatively.11(a)J 2. Using F : 744'. Using equation (3.0. 3.6L2j7.3456 From equation (3.52924.10) we can compute the coordinates of the center of the projection circle of plane p: C*: 0j2924 C' : -0'72844 We can draw the projection circle by three methods: 1. . 3 From equation (3. Stereographic projestion of a plane from its Normal Given the coordinates of the unit normal vector fi of plane p.13) with R:1 X : O.86 Graphical Methods: Stereographic Projection Chap.9) we compute the radius of the projection circle of plane p: r :1.8) the distance from the origin (O) to the center (C) of the projection circle of plane P is OC :0. and OC: 0. Cr: : -0.4t42 The distance from origin o to the center C of the projection circle is OC:l The coordinates of the center C of the projection circle of plane p are C* :0'86602 C' :0'50000 We can now draw the projection circle by any of the three methods of Example 3. using equations (3.70710 the radius of the projection circle of plane p is r : 1.4.. Using F : lM" and r : t.3456 [Fig. 0.:0.90040 tFie.3456.3.

l9l7 0 6 2. c* cy I 1.68223. and the degree step is 10'. For R : 1.9238 +2.4705.516.7474 0 8 5.1524.0o. -1).15.7928.2r3.Examples Example 3. r : I.23) with d: lO". from (3. Example 3. Entering R : 1.22) as C" : L. (3.18) gives the distance from the center (O) of the reference sphere to the center of the small circle (C) as OC : 1. 3. k: I.064r +o. Example 3.6. R 1. Constructing a Stereographic Projection Net (Equal-Angle Projection Net) In this example we compute the centers and radii of all the circles of the stereonet as shown in Fig. Alternatively. The unit vector corresponding to v is 0 : (0. and S 20".7320 0 7 2.7.0401 and from (3.70710).3734 and C" :1. the values of the radius r and coordinates C.s7735 0 4 1.70710 in equations (3.7r9 we obtain the radius r and coordinates C* and.36397 0 3 1.0. Draw a Small Circle Representing a Gone about a Vector v Given v : : = (0. and (3. -0. of the centers of the small circles are: .0154 +o. 0.6712 0 Similarly.0000 +r. and the cone makes an angle 6 : 20" with . 1.22) gives C* :0 and C":3. and Z : -0. Draw a Small Gircle Representing the Stereographic Projection of a Cone about the Normal to a Plane Given fi is the normal to a plane with u: 1-16" and P: 50". and C.70710.3054 +0.5.1547 ltro. Using (3. the coordinates of C arc determined by equations (3.16) determines the radius of the small circle &s r : 0.. X : 0.r7632 0 2 t.7587 +s. Y : 0. Assume that the radius of the reference circle R : 1.24).70710.20).4. with d : 1. Cy of the great circles.83910 0 5 l 5557 tL.

(Jn. Lr.31520 0. Graphical Methods: Stereographic Projection Chap. Ur. Pr.5557 6 1. . Pr.57735 0 tr.7474 0 +2. Using (3.7587 5. Ur. 3 cx cy I o.3054 5 1. .6712 0 P+ 1. (J n. Region 2 is the joint pyramid Ur. Uz. Ln. and Pn. we compute the radius r and the coordinates C.9) and (3.59239 r.17(a). Ln. Stereographic Projestion of a Joint Pyramid Given four sets of joints: Dip. U".00m 7 2.18198 Then construct all projection circles as shown in Fig.6n6 Pz 1. Region 3 is the joint pyramid Ur.9238 8 5.rs47 4 0.36397 0 *1. Each region is the projection of a joint pyramid. For example: . p (dee) (dee) Pr @ 70 P2 50 zffi Pt 80 0 P+ 20 150 Assume that the radius of the reference circle is R : 1. . a Dip Direction. cy Pt 2 o. Plane C. Ur. 3. Ln. .19t7 0 +1. Region I is the joint pyramid U r.67t2 0 +5. Lr. and Cy for each projection circle of planes Pr. Lr. Us. Region 14 is the joint pyramid Lr.0641 -0. Region 10 is the joint pyramid Lr.5557 -o.10).2Q694 -1.7320 0 +2.7587 Example 3.83910 0 *1. Lr. .8.0154 2 0.17632 0 +1.t736 Ps 5.0641 3 0.

0. Measure the distance AB.2.82303) Example 3. compute d: ON :0. .2 and Yo : 0.99.64. Now plot C and draw plane P as the circle of radius r about C.10.2. formulas (3.29).3.2.LI43 and OC: 0.88. the position of the projection N of the normal to plane P is now known.18(b).2.5322) Similarly.64 in (3. Example 3.0.34) determine a vector N - (X. The radius is then AB -T: t. equations (3. Y.4918. Using (3. Finding the Plane (P) Normal to a Given Vector in the Stereographic Proiection Given a poiqt i/ in the stereographic projection plane as shown in Fig.23260.11. 3.3.31) give r : l.3a) defining vector b corresponding to point B: b : (_2.3) B: (_2. line AB is the diameter of the circle of P along the line through the origin (O). and (/.30) and (3.18(b) Suppose that R : 1.B input Xo : -2.0.38) give the coordinates of the center (C) of the projection circle for the plane common to a and b: c: (_0.tl44 Assume that R : 1. for point . Finding a Normal of a Given Plane P in the Projection Plane Given a projection circle of plane P as shown in Fig.1249) Then equations (3.64) as shown in Fig. Let Xo: -0.Z) corresponding to point l: a : (-0. Considering N as the projection of a vector n. Our task is to draw a great circle through points A and B.18. _2.9.Examples 89 where Z. 3.23260. means the lower half-space of plane P. A Plane Gommon to Two Lines Given two points in the stereographic projection with R : 1: A: (_0. means the upper half- space of plane Pr. -2.72639. Example 3.88 and Yo : 2. 3. We can measure d: ON :0.

0. we find inter- section points Q and Q'.5. and Z.39) the angle::Xn vectors V. Using formulas (3.11 162. Xz. (d sin p. Then from (3. and v. 3.5) yz : (-0.35) to cal- culate Z. Finally. v : (_0. The vector v is represented by V: (-0.12. and vr. to unit vectors 0. and V"is Example 3. and V. and Srby dividing each component by the length.5. Yr. 0. Measuring the Angle between Two Vectors Given two points Vt: (1.. Stereographic Projection of the Orthographic Projection of a Vector v on a Plane Given plane P with normal fin and vector v. Let R : 1.42.Yr): (-0.37) . substituting Xr.15 and F :205" From (3. 0) in the stereographic projection plane. compute d: ON : 0. Let R : 1 and let the stereographic projection of fio be N : (-0. The construction is shown in Fig.24.0. Yr. use equations (3. d cos f) : (-0. Cr) : (1. 1. which are the orthographic projections of v on plane P.6973b). -0.42. (C. N.-0. Constructing this circle.0. Measuring the Angle between a Vector v and a Plane P See Fig.29. Zr. 3.4805. Example 3.15). and Zr. in (3.38) calculate the corrdinates (C") and (Cr) of the center of the great circle through V and.21(a). -0. Given a projection circle of plane P. 0.44). the coordinates of vectors V.34). measure the radius r and the dip direction f from this circle: Suppose that r : 1.Yr): (-0.0. The problem is to find the stereographic projection of v in plane P. are deter- mined as follows (with R : 1): Vl : (1.29).23938) Given the projection I/ of vector v.15) and (Xr. Graphical Methods: Stereographic Projection Chap. -0. 1) and Vz: (-0.44). which are the projections of vectors v. 3 Example 3.20.26413 The coordinates of N are nf . Then with (Xr.24.14.375) Convert v.13.

3. which have projections 'n/ and % respectively.86957) 0: (-0.60606. fz:0" From (3. comPute < (fr'D) :67'500 The angle between P and V is 90' 67.55.22(a). 0. of P. and Pr.46442 N: (dsin p. the dip direction F :210". 0.5" :22.76404. This could have been read directly as shown.21(b). and N. Measure the radius and dip direction from these two circles: Prt rr : l'!5. and Pr. d. fi: (-0. 3. -0.44753' 0'86957) fiz : (0.64516) Using (3. 47 502.46442) From (3. Example 3. -0'23938) Nz : (d. 0. cos B. Fr: 205o Pzi rz:1. d.33). we have 4 (fit' fi") : 77 '345" which is the d angle between P. and P"as shown in Fig. l[r : (dr sin Fr. compute unit Yectors fi and D.33). Finding the Direction of Sliding on a Single Face SeeFig. 91 Examples Using (3. -0'40220) . dcos P): (-0.) : (-0'11162. sin Fr.29). and fi. cos Pr) : (0. Measure the radius and dip direction of P. Given a projection circle of plane P and projection f of resultant force R.20868.5" - Example 3.15. respectively. From (3.16. Measuring the Angle between Two Planes Let R:1.39).29). 0.0. are the projections of normal vectors fi. comPute dr: ONr:0'264t3 dz: ON2: 0'46442 where l/. -0.63800) From (3. 0. compute d: ON :0.M753. compute fir : (-0. Given the projection circles of planes P.20868.55. The radius r : 1. Let R : 1.39).23221.

and Pr. 0). Calculate d:Or:ffi:O. From (2.0.0) is a point in the stereographic projection of plane P .9711. compute the radius ro and center Co of the plane normal to f : /o : 1.1333 OCo :0'5333 Draw circle Po.13) Using (3. 3 Measure the coordinates of f in the stereographic projection f: (_0. fo. _0.25 Using (3. The intersection point I* and i lie on the same side of circle Po.24.f12 is the projection of the sliding vector.Sr and 52. with dip a and dip direction p (with sign conventions as in Chapter 2).30) and (3. THE STEREOGRAPHIC PROJECTIOIU OF A PLAIUE IS A CIRCLE Given plane P. and given the projection r of resultant force r. APPETUDIX IMPORTAIUT PROPERTIES OF THE STEREOGRAPHIC PROJECTIOil t. Let R : 1.7) the equation of P is sin a sin pX * sin d cos pY + cos uZ : O (1) Suppose that (Xo.30) and (3. Note that points and r are on ^9.35) and (3. Example 3. respectively.31) with d: Or : O. Given the projection circles of planes P. so .58983 Draw circle Po with radius ro and center Co.23(a).25 compute the radius ro and distance OCo of the projection circle of the plane (Po) normal to vector f . passing through (0. 0. Measure Or:0. 3. . Graphical Methods: Stereographic Projection Chap.21294 Then using (3. the same side of circle Po . therefore.1609 and OCo :0.S1 is the projection of the sliding direction.17. compute the center C of the circle through points N and f: g : (I. ro : 1.10525) Then draw this circle.31). which intersects circle P at points . Finding the Direction of $llding on Two Faaes Simultaneously See Fig.38).6) and Q.

0) on d: the angle between d and any straight line on the cone is f (Fig.:$qffi-p.24). f)' :(#J' Rtandcos (7) Equation (7) defines a circle with center at (Xr' %): (R tanu sin B.0) is given bY X: Xot Y:Yot Q) Z:R(t-1) The piercing point.25).24 Stereographic projection. The focus of the projection is at (0. from Q). of this line on the reference sphere is defined by a value of t : to. XA . of the line on the sphere are. into (3) gives zRz ts: YTW#T (4) So the coordinates of the piercing point. 0. R tan d cos B) and with radius -R/cos c. Rtan asin f)' +(% . I. obtained by equating the line and the reference sphere Xz+yz+Zz:Rz (3) Substitutin1 (2).Zo). it satisfies (1). 3. 2.0. 3.0. -R) and a straight line from the focus to (Xo.fA):0 (6) which simplifies to (X. with lo in place of /. The equation of the cone is . -R) Figure 3. THE STEREOGRAPHTC PROJECTION OF A CONE ls A clRctE Given unit vector d:(X*Yo. tmportant Properties of the Stereographic Projection 93 (0.'n#le'ffi) (s) Since 1is also on planeP. combining (5) and (1) yields (sinasinp 2R2)Xo*(sindcos P2R2)Yo*cosdR(Rz . . . which is the axis of a cone with vertex (0. (Fig. %. I.

0. 3.)' * (r' . : _4 cos@ gio d:_ . Then. 2R2Y"Y. Y. 3 (X..Y):(#-t.Iis also required to be on a cone through 0.r. the radius of the circle is 4sinr0 '.Z (B) x?+Y:+Z?:l (9) As in the preceding section. Zo r-et a be the normal vector to plane p. ffi. with dip a and dip direction p.##) (14) . JXry-We cos d : XoX + Y.)' Equation (11) defines a circle with center at : (4_d-.7). (5) can be equated to (8).O) be a projection point of plane p and let the focus be located at (0.)' (l 1) (xo. This gives Rcosd:7ffip+ffffi*ffi 2R2X"X.#. Yt +:w (10) which simplifies to (x. -R) so that a straight line from the focus through the projection point intersects the reference sphere at I (Fig.0) Figure 3. Graphical Methods: Stereographic Projection Chap. from (2. let (xo.25 A cone about line ri.#r") and .):(#F+#. yo. The coordinates of I are given by (5). Xo: sin a sin B Yo: sin a cos p (r2) and Zo: cos d Substituting (12) in (11).zl 0 a (0. If point . .0. cos d+ cos d (l 3) and the coordinates of the center of the circle are (x.Y + Z.24).

x:0 and forPrr trz. lmportant Properties of the Stereographic Projection 3.X:0 The planes are represented. by great circles Pt and pz Fig.App.26).i trr. THE STEREOGRAPHIC PROJECTIOTU PRESERVES THE ANGTES BETWEEN PTANES *The stereographic projection is an equal angle This property can be stated: projection. and P" lEie.:6?+Y?+Z!)trz and i: 1. Z2) and suppose that Zt ) 0 and Zr2 0. 3. in the stereographic projection.26(a)] is exactly equal to the angle d2 between the tangent vectors to great circles P. Let (xr. and nr.Yt.r'?) where R. If the proposition is true. the angle d. Let Ir: (Xr. and P2 are two planes defined by normals n. between planes Pt (a) (bl Figure 3.Yr." Suppose that P.3.Z) and D2 : (X2rY2. and.26 Stereographic projection of the angle between planes.26(b)] at their points of intersection.2 . zt): (#r'f. P. [Fig. respectively: forP. 3.

96 Graphical Methods: Stereographic Projection Chap. 3

Then x?+y?+z?:l, z,)0 (1S)
suppose that the radius of the reference circle is unity (R : 1.) Let (A,, B,) be
the coordinates of the center of great circle P, and let its radius be rr. Using

A,'21 -xt-

+ (16)

and-l ,r: z
the equation of circle p, is
(x - A,), * (Y - B,), : r? (L7)
Solving equation (ll) for Y,
y:+m*n, (18)
Differentiating with respect to X, the tangent vector (t ) to the great circle for
plane P, is

t':(l \ (re)
t, is parallel to
(m,+(x-A)) (20)
From equation (18),

so from (20),
tr : [*(F - B,), +(X
- A,)l
or t, : +[(f - B,), -(X - A,)l el)
From (17), lt,l : r, and accordingly,

f t' : *lft" - B,),
-(x - A,)l (22)
- tt'
Let (x, Y) : (xo, ro) be the point of intersection of circles p, and p, so that
(&, %) satisfies (Z2)withboth i: ! and i :2. At(xo, %), the angle between
f, and f, is given by
cos d, - f r.f z: I#n*, - Ar)(Xo
- Ar)+ (% - s,Xro - Br)l

+;K&- A,)'*(Yo-8,)'
+ (& - Ar)(A,
- Ar)+ (% - Br)(8,
- Br)l
: + *rn,+ (& - A,)(A, - A;)
+ (% - 8,)(^B,
- Br)l

lmportant Properties of the Stereographic Projection

This yields
cosd2: f[x,xr*yry, I zrzz* Xs(xiz- xzzr) * Yo(ytzz-!zzr)l Q4)
The line of intersection of Pr and P, is found from
fir x fiz: (Itr) : (Xrr,Yrr, Zrr) (2s)
Irz: (yrr, - !zzr, xzzr - xrzz, xr!z - xz!) Q6)
The stereographic projection of I,, is (Xo, Io); using (3.5)'

(xo,ro):(+,+) (27)

where k- l\zli- xtYr- xrYr

Substituting for X and Y from (26),


Consider the term
Xo(xr 22 - xzzt) * Ys(ytzz - fzzt)
which is found in Q$. Substituting for Xo and Yo from (28), this term becomes

P r*)(x,r, - xzz t) * W)(y,r, - !zz)
which is equal to zero. Therefore, (24) simplifies to
cos: f [xrx, * yry, * zrzzl
d, Qe)
But lxrxr*yJz* ztzzf :frt,fir: cosdl
So d,- d1 and the angle between the vectors fi, and fir,hence between Pt and
Pr, is exactly equal to dr. Q.E.D.

chapter 4

The Removability
of Blocks

The central idea of this book is that stability analysis of excavations can skip
over many of the conceivable combinations of joints and proceed directly to
consider certain critical blocks, denoted as keyblocks. This efficient approach
is workable by virtue of a theorem on the finiteness of blocks. tn this chapter
we establish and demonstrate this theorem and some associated propositions.


You will recall from Chapter 2 that a block is determined by the intersection of
a particular set of n half-spaces. Considering orientations alone, there are 2,
unique half-space intersections. Not all of these define potentially critical blocks.
We will now establish criteria for the relative importance of blocks. A keyblock
is potentially critical to the stability of an excavation because by definition, it is
finite, removable, and potentially unstable. Table 4.1 uses these terms to recognize
five types of blocks. An infinite block (type v), as depicted in Fig. 4.1(a), pronid.t
no hazard to an excavation as long as it is incapable of internal cracking; we
have ruled out consideration of internal cracking of blocks in the basic ur*tttp-
tions introduced in Chapter 1. As we shall see, most of the 2' half-space intei-
sections produce infinite blocks. Finite blocks are divisible into nonremovable
and removable types. An infinite block that does not crack obviously cannot
be removed from the rock mass. However, a finite block may also be non-
removable because of its tapered shape. We shall prove later than any finite
block intersected by an excavation surface so as to increase the totat number of


Types of Blocks


aPEffi 8rc95!

lnfinite F inite

Nonremovable Removable

IV l\
ra'ered -/
rr/ I \,
Stable aren without Stable with Unstabld without
friction zufficient support

Potential Key block

Key block

faces acquires a tapered shape and cannot be removed from the rock mass.
An example of a tapered block (denoted type IV) is presented in Fig. a.l(b).
All tapered blocks are nonremovable unless they are undermined by movement
of an adjacent block.
Nontapered, finite blocks are removable but they are not all critical to
the survival of a given excavation under a given set of loading conditions. We
can distinguish three classes of removable blocks. A type III block has a favor-
able orientation with respect to the resultant force, so that it tends to remain
stable even without mobilizing friction on its faces. Figure a.1(c) presents an
example. Although it would be possible to lift this block from its home, the
block is of no concern under purely gravity loading since its virtual movement
is away from the excavated space. Figure 4.1(d) shows a type II block; it is
defined as a block that is potentially unstable (since the tendency for movement
is toward the free space) but unlikely to become unstable unless the frictional
resistance on the potentially sliding face is extremely small, or there are loads
in addition to the block's self-weight driving the displacement. A block like this
is a potential key block. A true key block (type I) like that of Fig' 4.1(e) is not
only removable but oriented in an unsafe manner so that it is likely to move
unless restraint is provided. In the case shown in Fig. 4.1(e), the restraint would
have to be constructed before the excavation has completely isolated the block.
The actual shape of a rock block is governed not only by the numbers and
relative orientations of the planes forming its faces but by the numbers of these
planes that are paired. A block may be formed with one each of n discontinuities.
br it may have one or more sets of discontinuities represented twice as opposing

100 The Removability of Blocks Chap, 4

(a) (b)



Figure 4.1 Types of blocks: (a) infinite; (b) tapered; (c) stabre; (d) potential
keyblock ; (e) keyblock.

faces. Figure 4.2 illustrates this with a two-dimensional example. The block in
Fig.4.2(a) is determined by the intersection of three hatf-planes, defined by three
nonparallel joints. Repeating one of these joints creates a block like that of Fig.
4.2(b). Repeating all the joints outlines a block like that of Fig. 4.2(c), which
begins to resemble a crystal. Blocks with parallel faces tend to be more stable
than those lacking parallel faces because the range of directions of kinematically
possible movement become diminished.
Two-dimensional tapered blocks are shown in Fig. 4.3. In Fig. 4.3(a) a
triangular block is to be cut by the excavation surface shown. The resulting
four-sided block will subsequently have a shape such that it cannot move into

Theorem of Finiteness 101

(a) (b)

Figure 4.2 The influence of the number of parallel sides on block shape: (a) no
parallel sides; (b) one set of parallel sides; (c) all sides parallel.

the excavated space; that is, any displacement of the block toward a point inside
the excavation would generate a larger width so that the block cannot fit through
the available space. This is also true of the block produced by four nonparallel
joints in Fig. 4.3(b). When the block is excavated as shown, the resulting flve-
sided block becomes tapered. If the excavation surface eliminates more than
one corner of the originat block, so that the resulting number of faces is not
increased, the resulting block may not be tapered. This is depicted in Fig. 4.3(c),
where two corners of the original block have been removed, and in Fig. 4.3(e),
where three corners have been excavated. Neither of the resulting blocks are
tapered. These examples are two dimensional but they apply equally to three-
dimensional blocks.


An important theorem will allow us to establish whether or not a given block is
finite. Let a block be defined by the intersections of half-spaces defined by
planes L,2, . . . , n. Using the terminology introduced in Chapter 2, we defrne
an associ ated block pyramid by moving each plane to pass through the origin.

A convex block is finite if its block pyramid is empty. Conversely, a convex block
is infinite if its block pyramid is not empty.

It will be recalled that a block pyramid is generated from the system of
inequalities for a block by setting all terms D, to zero, as illustrated in Example
2.9. An'oempty" pyramid is one that has no edges. In Example 2.9 we saw
how to determine whether or not a block pyramid is empty through the use
of vector analysis. It is also possible to judge the emptiness of a block pyramid
using the stereographic projection. Before proceeding to demonstrate how to do
this, it will be instructive to pursue some two-dimensional examples.

102 The Removability of Blocks Chap. 4


Figure 4.3 Two-dimensional tapered blocks.

(al t2l Region common to u?. 4. Since there is a region common to these three half-spaces. Lr. and L3.Theorem of Finiteness 103 Figure a. The block pyramid cor- responds to the region common to (Jl.) The block is obviously infinite.a@) shows a free surface (3) and two joint planes. two- dimensional example. (The symbols L and (J are used. To apply the theorem. as in Chapter 2. to identify the half-spaces respectively below and above a given plane. block UtLzLiis determined to be infinite. Ll.and L! u! (b) Figure 4. This has been achieved in Fig. L8.4 Application of the finiteness theorem to an infinite block. . we determine the block pyramid corresponding to UtLzL3 by moving the half- spaces. andZ! on this shifted diagram. without rotation. Consider the block determined by the intersection of half-spaces Ur. as required to pass the bounding planes through a common point.4(b).

5 Application of the finiteness theorem to a finite block.5. The method is particularly simple when the stereo- graphic projection is used.104 The Removability of Blocks Chap. [n real. the finiteness of the blocks considered was obvious by inspection. Block (/1(JzL3 in Fig." By the theorem. .5(a) is finite by inspection. However. To determine this formally using the theorem. In these two-dimensional examples. the common part of Ul. inspection is not so simple. planes I. 4 The converse is shown in Fig. There is accordingly no edge to the block pyramid UoJl\L?.2. which is therefore "empty. Ug and L! (b) Figure 4. Uorhas no points overlappingregionZ! except for the origin itself. as will be shown. three-dimensional cases. Free surface An "irclated block". and 3 are shifted without rotation to pass through a common origin. the formal application of the theorem does permit a direct determination of finiteness. It is then seen that the region common to U! and. two-dimen- sional example. Ls QI (al The only point common to U?. block (Jr(JzL3 is finite. U2. 4. a.

Theorem of Finiteness 105 In each example. and U! is the space pyramid SP. Figpre 4. that is.6(a) shows the projection of the four planes using a lower- TAB. Each block pyramid was therefore formed of planes parallel to both joints and free surfaces.5(b). As another example. a Dip Direction. Since JP is not included within SP. the block pyramid must be empty.a(b): U\LZ defines the joint pyramid JP: Lt is the excavation pyramid EP. In Fig. The block pyramid (BP) is then the intersection (n) of the joint pyramid and the excavation pyramid for a particular block: BP:JPNEP (4. Since JP is entirely included in SP.3) Let us now reexamine the previous examples in Fig.2. A block is finite if and only if JPnEP:@ (4.8 [Fig. as the set of directions that is complementary to EP. 3.2 Dip.2) is equivalent to stating that a block is finite if and only if its joint pyramid is entirely contained in the space pyramid. it is possible to produce a stereographic projection of the series of planes by placing each in position to pass through the center of a reference sphere of radius R. a. the block is infinite. The set of shifted excavation half-spaces will be designated as the excavation pyramid (EP). The Finiteness Theorem on the Stereographic Proiection Given a series of joint planes and free surfaces.LE 4. the block was defined partly by joint-plane half-spaces and partly by free-surface half-spaces. The joint-plane subset of the half-spaces determining a block pyramid will be designated as a ioint pyramid (JP). Each shifted plane projects as a circle crossing the reference circle at opposite ends of a diameter. 4. that is. p (dee) (deg) Joint plane I 30 90 Joint plane 2 65 50 Joint plane 3 65 130 Free plane (excavation) 15 90 . A typical projection of the great circles for a series of planes was demonstrated in Example 3. the block is finite.2) An alternative statement is possible if we define a space pyramid. JP is Uor(lL and SP is U3. if and only if JPcSP (4. SP: -EP Then equation (a. we will construct the stereographic projection for blocks defined by the three joint planes and single excavation surface listed in Table 4.1) For a block to be finite. SP.17(a)].

4 Reference circle \ I 111.6 Application of the finiteness theorem in three dimensions.106 The Removability of Blocks Chap. . t0 I : \ \ (a) (b) Figure 4.

. consider only the joint planes. The free plane is represented by the dashed circle and the three joint planes are represented by the three circles with solid lines.Theorem of Finiteness 4: Free plane {d} Figure 4. In this lower-focal-point (upper-hemisphere) projection.U!. The number 0 corresponds to the symbol U and defines the half-space above a plane. Thus 100 signifies the joint pyramid LorU\. The number L corresponds to the symbol Z and identifies the half-space below a plane. Each is identified by a string of three binary digits. The regions of intersections of these three circles define joint pyramids. and above plane 3. The digits are arranged in order. above plane 2.6 (Continued) focal-point stereographic projection. For now. the region above a plane is the area inside its great circle. The reference circle (horizontal plane) is the dotted circle. which is simultaneously below plane 1.

THEOREM OtU THE REMOVABILITV OF A FINITE. partly inside EP) and are therefore not entirely con- tained in SP. That property is determined by a second theorem. We can determine at a glapce that 011 is the only joint pyramid entirely contained in the SP above the free plane. convex block is removable or not removable according to its shape relative to the excavation. A convex block is removable if its block pyramid is empty and its joint pyramid is not empty.6(d). they are. leaving only the two joint pyramids.e. cotuvEx BLocK A finite. Block 100 is shown both in this section and in Fig.108 The Removability of Blocks Ghap. joint pyramid 100 corresponds to a finite block- LrU"Us. All the other regions identified in Fig. nontapered. 011 and 100. a. by the finiteness theorem. an isometric view.. It is a type I block-finite. the free surface has been redrawn as a solid circle and portions of the joint great circles have been removed. By the finiteness theorem. When tie blocks are formed below the free plang. and probably unstable without support unless it has a considerable friction angle on the bounding planes. it is entirely contained in SP. the space pyramid becomes the region above the free plane and therefore inside the free-plane circle. that lie entirely inside or outside the free-plane circle. Therefore. determines a block that is finite below an underground opening that has the free plane as its floor. 4 Consider now the excavation of an underground opening having the indicated free plane as its roof.6@) shows a section through the roof looking south. 4. The space pyramid of the excavation with the free plane as its roof is the region below the free plane. we will see that region 011. SP is the region outside the free-plane circle. infinite. Blocks in the roof are formed of joint-plane half- spaces together with the half-space above the free plane. Figure a. Such a block is quite clearly a potentialhazard for the excavation. but it is not necessary to draw a block to determine that it is removable. With the same line of argument. which lies entirely inside the free-plane circle. Therefore. accordingly. The drawings of the block are helpful in planning rock reinforcement. Since joint pyramid 100 is entirely outside the excavation circle. A convex block is not removable (tapered) if its block pyramid is empty and its joint pyramid is also empty. We have previously termed a nonremovable finite block as tapered.6@) are at least partly inside the excavation circle (i. the block considered in both parts of the removability theorem must be finite. a. Necessary and sufficient conditions for the removability or nonremovability of a finite block are established by the following theorem.6(b). The new theorem states that a finite block determined by a series of joint planes and free surfaces will be tapered if the . In Fig.

is also empty. Us) tapered Region common to 1 uf.7(a) is determined by half-spaces LrUr/rZr. JP" is L?U\F. there is a region common to these shifted half-spaces..J A(U1. A two-dimensional example will help to demonstrate the theorem. On the otherhand. a3@) and (b). if the block corresponding to the joint pyramid is finite. and therefore JP. the block corresponding to the joint pyramid plus one or more free surfaces will be tapered. that is. Convex Block 109 joint-plane half-spaces themselves determine a finite block. a. L {!ock B in Fig. 4.7@). by r-"*ffi u+ 4 . Ua) not tapered B(L1. Therefore.IP" is empty. u! L2 u4 The only point common to L?. ug.Theorem on the Removability of a Finite. and L! (bl (c) Figure 4. is not empty. u!. Figure 4. . BP. determined by UrUrUnLr.The only point common to these shifted half-spaces is the origin.7 Application of the removability theorem in two dimensions. as showir in Fig.?. By the first part of the theorem. is not empty and BP.7 shows a series of blocks defined by joint planes and a free surface (plane 5). We saw this pre- viously in Fig. JP for this block is U? UZU|. Consider first block l. block I is removable because JP. As shown in Fig. a. U2.7(b). is empty. U2. Therefore.theblockpyramid UiUgU|L? ls empty.

its joint pyramid is empty. block -B is nonremovable. An empty joint pyramid is one that has no edge. it cannot be represented on the stereographic projection. If a block is finite. except those that just happen to lie on the surface of the reference sphere.110 The Removability of Blocks Chap. p Joint Plane (dee) (dee) 1 70 l0 2 60 il0 3 Q 230 4 20 330 .4) The stereographic projection can represent lines and planes in space.) is given by l[r:n(n-l)+Z (4. the only point in common to the half-spaces defining an empty joint pyramid is the origin itself and the origin is absent from every stereographic projection except one of zero tadius-a trivial case. It will be proved in Chapter 5 that the number of regions actually appearing on the stereographic projection (lf. 4. Each of the joint pyramids is identified by a four-digit binary number.3. a Dip Direction. N^ : L4 and N" : 2. The regions that are absent from the stereographic projection are therefore the joint pyramids corresponding to finite blocks. When that block is defined entirely by joint half-spaces. it will TABLE 4. 4 the second part of the removability theorem.s) Consider the system of four joint planes listed in Table 4. Application of the Removability Theorem in Three Dimensions Using Stereographic Projection Recall that the joint pyramids belonging to a given block pyramid plot on the stereographic projection as a series of regions enclosed within portions of great circles. there are T possible blocks created by their intersections. The stereo- graphic projection of these joints is shown in Fig. However. The number (lft) of these blocks defined by n nonparallel joints is therefore i'rfr:2"-[n(n-1)+2] (4. The case of a nonconvex block is treated at the end of this chapter. Checking them off in turn.8 (a lower focal point projection). this means that it is tapered. Since it is finite. when n is greater than 3. its block pyramid is empty. In other words. Given n nonparallel joint planes. If it lacks an edge. Since n equals 4.3 Dip. But it cannot project points in space. not all these possible regions appear in the stereographic projection.

it cannot fall into the tunnel. where a circular tunnel has a finite block in the roof. Tapered blocks are important to excavations because they can not be key blocks. be confirmed that only 14 regions appear in the projection plane.9. 4. The significance of finite blocks defined only by joint planes is explored in Chapter 5. .8 Lower-focal-point stereographic projection of the joints given in Table 4. Since this block is tapered. The two finite blocks determined by this system of joint planes are therefore (J. The two that are missing are 0001 and 1110. An example is shown in Fig. When either of these blocks is cut by a free surface to define a new block whose block pyramid is the appropriate joint pyramid (0001 or 1110) plus the free surface. the resulting block will be tapered and nonremovable from the rock mass.(J2(J3L4 and LtLzLr(Jo.3. Convex Block Figure 4.Theorem on the Removability of a Finite.

. There is. ja+r has an empty joint pyramid. SYMMETRY OF BLOCK TYPES In the preceding examples.6) :::: A^X + B^Y + C^Z: D^ . Since any tapered block i. 4 Figure 4. il*r is finite. in fact.3 that two blocks were tapered: 0001 and 1110. block 100 was finite in the roof above a given free plane.X * Bzr.1r2 The Removability of Blocks Chap. Thus the cousin of any tapered block is also tapered. The block whose joint pyramid is tttttg . Blocks that are related as in the conditions for the proposition will be termed cousins.. i2** is finite if and only if the block whose joint pyramid is jli|jg .. . The symbol is i I j I either or 0.* Dz "r7: (4. For example.6. while block 011 was finite in the floor below the same free plane. the symmetry proposition establishes that the corresponding symmetrical block j. including a mixture of joints and free surfaces. The symbol is if i is 0 or 0 if i is I. . .9 Tapered block above a tunnel. .. j9 and whose excavation pyramid rb j9* r j|+z .. a general proposition concerning these observations: Consider the convex block formed of n joint planes and k free surfaces. has an empty joint pyramid.. The equations of these planes are ArX * BrY + Cp: Dr A". . PROOFS OF THEOREMS AIID FURTHER DISCUSSIOil Finiteness lheorem.. we deter- mined with respect to Table 4. In the example shown in Fig. 4. j. Let there be m planes. . iz . you have no doubt noticed a certain symmetry in the solutions to the removability and finiteness criteria. tf and whose excavation pyramid is il*i}*z .

v.9) where I is a constant varying from 0 to I in moving from I to L If all I values of between 0 and 1 satisfy the inequality (4. In a convex block. both within the half-space: ArX * BtY + CLZ) D. (4.. Fig. * (1 .Y. Figure a. l)Y i.Y. x)xi. However. Zt) and J : (X1.I)zi (4.8). Consider two points. * (1 .Proofs of Theorems and Further Discussion Consider the following system of half-spaces: ArX * BtY + C&) D.10 Convexity of a block: (a) aconvex block. A convex block is defined as a region within the intersection of a number of half-spaces such that a line between any two points in the region is contained completely within the block.Y + C|Z2 Dz (4. (b) aconcave (nonconvex) block.10(b) shows a block that is not convex ("concave'n). This can be seen intuitively. can be positive or negative.9) ..I and J is convex. 1: (X. -l (1 . all interior angles are less than or equal to 180'. then by definition the region between .Iand J are permitted to be anywhere in the half-space. u:* * Bz. 1. since the con- stants A.x. 4. and D. Z1).10(a) shows a convex block.. A half-space is an unbounded convex block.Y1.1) A-X + B*Y + C^z> D- There is no loss in generality in adopting > for all inequalities. it can be established formally as follows.8) Any point on a straight line between 1 and J has coordinates given by linear interpolation as v. Substituting (a. The entire half-space is convex if . FigUre 4.llz. Br C..

Yr. Since any two points in the intersection are in both blocks and each block is convex.7) is a convex block. tYo.11b). The "intersection" means the part of each block that is common to both. and NlD. tzo). the intersection of all rows of (4. Therefore. Ar. the line between these points is within each block and therefore is in the intersection. Suppose that there is a point (Xr.lLa) and another point (Xo. and so on. According to this lemma. Since a half-space is a convex block.X*Bz.*CrZ. the intersection of any two half-spaces is a convex block by (2). The intersection of any number (m) of half-spaces defines a convex block. then the convex block ArX * BrY * CrZ> O A"X * BrY + C2Z> O (4.Y+CrZ2D. by (1).z (4. The Removability of Blocks Chap. Z) not equal to (0.. 0) that satisfies (4.11b) A*X + B^Y + C^Z> 0 represents a series of z inequalities in three unknowns with only one root- X: Y:Z:0. with t ) 0. Thus the intersection is convex. If the convex block ArX*BtY+CAlD. Zr) satisfying (4. 2. Then (tXo.1)N> Dl (4. 4 in the left-hand side of (4. they each satisfy (4. IM + (1 .8). 0. also satisfies (4.8) gives LM+(1 -r)r/ where tuf-Arx. if 1is between 0 and 1. SO MlD. Its intersection with a third half- space is then convex. 3. 4.*BrY. Proof of (4).1lb) because . and N:ArXt*BtYt+CAt Since by design both I and J are in the half-space.1la) A^X + B*Y + C^Z> D^ is finite. until all m half-spaces have been inter- sected. Yo. The intersection of any two convex blocks is also e convex block.10) and the block is convex.

. * tXJ.".F.(A. If the convex block given by the m simultaneous inequalities of equation (4. 0. .for A.Zo) =a 0 can not satisfy (4. + Cf)) D. But then (4. ^ ^:. . (Zt * tZ) satisfies (4. . (0..m. for all i:1. Proof of (5). A.0). Yo.2. 0) satisfies (a.B formed by n joint planes and k free planes.'. for i: 7r2. 0. .(E. 5. + Bfiz + CEs) foranyt>0 Then. But we supposed that only (0. :.11a) for any t> i:1. (Fr. . Blocks. 0. .?.(Er + Frt)) D.11a) is not finite. * B. Therefore.l )D.r2a) A.m. Z: Es* Ftt with (tr'r .Proofs of Theorems and Further Discussion tArXo *tBr%+tCrZo)0 tAr. meaning that X: Et * Frt.. Fr) satisfies (4.11b). Fr) ?€ (0.Y + C.. * Frt) * C. Fr.F.11a) for any positive value of r. then it has at least one infinite edge.Fz + C[3 cannot be ( 0. Y: Ez* Fzt.*0 then (X.' * c. therefore..X * BrY + C&) D. Suppose that the block determined by the simultaneous solution of the inequalities (4..11b). Therefore. . there is a contradiction and (4.11a) is finite.z. ( yr + t fo). The inequalities of the block are A.2.:X' :.. u'. then the convex block given by the m simultaneous inequalities of (4.X + B.:'l:.11b) has only one root. 0. 0) satisfies (4. .11a) is infinite. * Frt)* B.* (4.Z> D.1i.1la) is finite. and t(AFt * B.E. . Block Pyramids.11b). (Xo. and Associated Regions Assume that there is a block ..(8. Fr..ffi.Xo * tBz% + tC2Zo > O tA-Xo + Since. Therefore.i'l:.

13a) A"X + B"Y + C"Z> O An*rX I B.t=0.Y+C&>O u:.12.**u'.*rY * Co**Z) Dn+t The system of inequalities (4.*rY I Cn+tZ> O (4'13b) :::: An**X * Bn*Y * C.'*':. The Removability of Blocks Chap. they define what we will designate as a joint block. The system of inequalities (4. 4.11 shows a typical block. . The rock mass is determined by the system of inequalities (4. (a.13a) determines the joint pyramid (JP) of B.*rZ> 0 The system of inequalities (4.13b) determinesthe excavation pyramid (EP) of B. Free plane i=3 Joint plane i=l Joint plane i=2 Figure 4. the common points of these half-spaces.12). in two dimensions.12a) are for the half-spaces cut by n joint planes.. with n :2 and k :2. Figure 4. The rock mass is then the intersection of the last (ft) half- spaces of (4. while the system (4. 4 An*tX * B. The system of inequalities determining the block pyramid (BP) of A is ArX*B.L3) defines the block pyramid (BP). that is. 1.12b) are for the half-spaces cut by excavation planes.12b) as shown in Fig.ll Block created by the Free plane intersection of two joint half-spaces and i= 4 two free-plane half-spaces. The intersection of the entire system (4.

.14./\ Rock Figure 4..12b) as shown in Fig. i=4 2. is the union of a block (B) and the free space.Proofs of Theorems and Further Discussion Free plane i=3 \ \// . 4.12 Rock mass created by the intersection of two free-plane half. The removable spece. that is.. Free plane spaces.13 Nonconvex rock mass created by the union of two free-plane half-spaces. . The free space consists of all points lying outside the rock mass. The union of ft half-spaces is the set of points in any one of them.n+2. A.13. It is therefore the union of (a) the intersection of inequalities (4..12). The free space is the union of all half-spaces i : n + t through n + k of inequalities (4. Fig.X*B.Y*CtZ<Dt (4.' .I2b) with ) replaced by <.n+k 3. Figure 4.r4) i:n*l. 4. and (b) the union of the (ft) inequalities of QJa).. The free space is therefore in the union of the half-spaces opposite to those of inequalities (4.

A finite block cut by joint planes and free planes is removable if it can be moved along a direction without colliding with the adjacent rock mass [Figure a.12al Figure 4. The intersection of all joint half-spaces of a block (the joint block) belongs to its removable space (Fig.14 "Removable space" of a block. The Removability of Blocks Chap.13a) is not empty. Removability Theorem 1.16(b)]. 4 Figure 4. an equivalent statement is that the block formed by joint planes and free surfaces together (4. 4. The necessary and sufficient condition for a block (8) to be removable is that BP is empty and JP (a.15). Space defined by inequalities 14. 4.15 Joint block belongs to the removable space of a block.16(a)]. . A finite block cut by joint planes and free planes is non- removqble (tapered) if it cannot be moved in any direction without collid- ing with the adjacent rock mass [Figure 4.12a) is infinite.12) is finite and the block formed by joint planes alone @. 2.

there is a direction (Xo. nwill not tend to close. be the normal to joint r pointing to the interior of the block (Fig. Yo.. Let n1 : (At.. Zo) : xo * (0. 8. -necessary . .2. .. that is.2.1s) xo: direction of movement Figure 4. C. .ft. 0) such that when the block moves along x0 every joint i : l.Proofs of Theorems and Further Discussion 119 /l I i I I I i i I -l\ xo: direction of Adjacent rock mass movement (a) (bl Figure 4.. .). . Proof of (2) The necessary condition: If a block is removable. nr. 0.. . i : 1..17 Removability of a finite block condition. 4.16 Test for removability of a finite block under sliding.n (4. xo >0 fot i:1.17)..

. Yr. i:1. Symmetry Theorem. (4'20) has only the solution (0.18) is finite.7.X*B..0.Y*CA<0...2. (0. A.' ) D. For any point (Xr.X*B. i:1. Therefore. . From paragraph 4 of the preceding section..21).) .n * k (4.X * B.Y * CtZ> D.16) is the joint pyramid of the block. Yo.Y * CIZ> O..16).0) is the only solution to the system A.lf i :..21) since if (Xo.:.. + B. Yo.z. (4.n for So the moved point (& + Xot.12a). * Yot. The Removability of Blocks Chap. 4 ArXo * BrYo * CtZo> O ArXo * Bzyo + C2Zo> 0 (4.20). -Zo) must satisfy (4. * c. After moving in direction xo.. If the convex block formed by the intersection of A. " '. Zr) become (X.. the intersection of all joint half-spaces belongs to the removable space.16) AoX..X * B. there is a vector Xo : (Xo. t > 0. The sfficient condition' If the JP is not empty.Zo) is a solution of (4. satisfies inequal- ities (4. (4.16). 0.. Zo) satisfies (4. (- Xo. Y. the coordinates of point (Xr. 0. Y. Yo. then Yo. the joint pyramid is not empty.2.'". * Zot)with t > 0. Yr. Z.19) is finite.18) is finite. the block is removable. * CtZt> Dt.0) such that (Xo. and (4. Thus the moved point does not lie within the rock mass outside the block.Y*CtZ<Db i: lr2. Proof.19) is finite. i:7.2.n + k. * Yot.0) is a solution to (4.2. then the intersection of the sys- tem of inequalitieq_ A..17) gives ^ ^! i.L2). Therefore..Z.fl*k. rtr * k. the convex block formed by the intersection of A.. * Zot). i:1.Yr + CZo) O Since (XoYoZ) * (O. the convex block (4.Z) * (0. :. 21) of the block defined by (4.Y. .0.1. Inserting this in the left-hand side of (a.0).:'. X:'l:. Assume fhat block (4. + Xot.Xr * B. Therefore..

of Bwhich points inside. . rn I k . UB(t') is bounded by face F. .. Figure 4. B is a united block. and includes f'... UB@) is a half-space. i: tr.. | !\ ^:. mB @. .n * k. of . In general..C.). is thenormal of free face F...n+ k The equation of ^F. .8.: (A.Shi's Theorem for the Removability of Nonconvex Blocks SHI'S THEOREM FOR THE REMOVABILITY oF tlomcoilvEx BLocKs United Blocks A united block is a union of convex blocks.B. . a united block is nonconvex.. is 4X + B'Y * C1Z: D' i: lr.fl.8. the number of convex blocks is flnite.i: n* 1. 0.18 shows a two- dimensional example of a nonconvex eight-sided block. The boundary of a united block consists of faces of joint planes or free planes.18 Decomposition of a united block.8.. Let 8...i: 1. Figure 4. Assume that . is the normal ofjoint face F.. . .Bwhich points toward the inside of ..

i:Ir. ^B(. The Removability of Blocks Chap. 4.:: (4) EP:UEP(r. 4 quation of UB(O. j:n* 1.such that rd e UB(0) and EP(/): n U@. Because .:n (4. In this example.. j : 1. are the points of B...23) where .n+k. . Ar.a) is a convex block.27) and (2) EPnJP: @ (4. we define a convex block B(A) as follows: B(A): /J uB@) (4.) is A... and is given by shi (1e82). then B(At)..) (4.n* k..rh (4. . One can prove an important fact that B(A) c. Assume that B is a united Block. .24) where F:the set of all j. B(At).2s) such that (a6n. Are B.E : the set of all. A. such that A e UB(0).h . j.. each of which is entirely within . such that I e UB(0). j:1.28) h where (3) JP:nJP(IJ (4.26) The necessary condition for B to be a finite removable block is (l) IP+ s (4. Then we have JP(/) : nu@. and A1.2e) . B (4. ... Denote by JP(...31) The proof depends on elemental theory of general topology.) (4.Y * C.) The equation UB(8.B is finite and B(At) c B.) (4. an eight-sided united block (B) has been decomposed into three convex subblocks B(At). .30) Proof.X * B.fl. .Z> Dt Suppose that A is a point of B.22) where D: the set of all j. The fact that B(A) c . Theorem on Removability of United Blocks.8.18.B is apparent from the two-dimensional case dia- grammed in Fig. . B(Ar). respectively... i:t.a) and EP(l) the joint pyramid and excavation pyramid of block B(A). .

. 4.) n JPI c [EP(11) n JP(.33) and (4. JP n EP : JP n t0 t=l gp(. h. IEP(.4.)l h : [J EP(.).4r) n JP (4. 4. . from 3.) n EP(. 4.34) we know that JPnEP:@ (see Fig.20)... Because B(A. 4... ..19 JP and EP for each component of the united block.h (4.B .): @.1.33) J=r Since JP : n JP(lj) h l=L JPcJP(lr).18).{ )l : a (4.h then. i:1.34) From (4.4. i : l. Figure 4. JP(A.32) (see Fig. then from the theorem on finiteness. is finite.I9).Shi's Theorem for the Removability of Nonconvex Blocks 123 is finite (see Fig. The above is shown in Fig.f) along which B can move without invading the rock outside of . i:1.20. .. This figure also shows that the set of all directions (.. .

4 I _* \ "/A I % I r!t Figure 4.4r) and there is a joint face Fo of B(Ar) such that s # u(60) So when .B at face Fo. see the compatibility theorem described in Shi (1981): . The Removability of Blocks Chap. such that s + JP(.B is moving along J. It can be seen that if i+JP:AJp(A) then there is an l. .20 Application of shi's theorem for a united block. For further detail.B will invade the rock outside of . is the set of all vectors of JP.

When blasting in a rock mass like that of Fig. 5. The properties of the rubble will depend at least partly on these same factors. general discussion in rock mechanics literature. and nonorthogonality of joints produce variable and nonprismatic block shapes. and arrangements of the in situ joint blocks.3(a)] . to our knowledge the subject of blocks within a rock mass has not received any significant. A type of rock block that is nonremovable from an excavation surface because of a tapered shape was discussed in Chapter 4. sizes. In this chapter we explore the geometrical properties and numbers of joint-block types produced by given joint systems. However. Joint blocks are the building blocks of a rock mass and must therefore be linked fundamentally to rock behavior. When. a single rock type may yield tabular blocks [Fig. without free surfaces. and there are no theories of rock behavior that rely on an accurate description ofjoint-block shapes. The design of a blast to achieve a particular shape of excavation must then be influence by the shapes. Such blocks do exist in the orthogon- ally jointed.2(b). 5. rock blocks have been depicted as building up a model rock mass. For example. as in Figs. behind the exposed face. only cubical or prismatic blocks were stacked. But more gen- erally some variation in extent and spacing. the joint blocks at least partly determine the sizes and shapes of the debris as idealized in Fig. 5. chaptet 5 Joint Blocks Joint blocks are rock blocks determined entirely by joint planes. uniformly dipping siltstones of western New York. 5. joint blocks exist inside a rock mass.1 and 5.2(a). in physical or numerical studies.2(a). In other words. that is. Delete the free surface from such a block and the remaining joint planes will determine a joint block.

t. IJtah.1 Joint blocks: (a) prismatic joint blocks exposed in the foundation excavation for Upper Stillwater Dam (USBR).'{ffi tun. * *'*'* . .ri*-"i Figure 5. (b) a section through joint blocks in the Upper Stillwater Dam foundation excavation. in Precambrian sandstone.S".

. I (a) (b) Figure 5.3 Block shapes produced by excavation depend mainly on joint block shapes: (a) a tabular block.2 Effect of blasting in jointed rock. 5 Joint Blocks Figure 5.Chap. (b) a cubical block.

5. There is an important difference between the joint blocks discussed in this chapter and the key blocks forming the main interest of all the other chapters. Intuition suggests certain principles on the relative occurrences of different .128 Joint Blocks Chao. because blocks with parallel faces have greatly restricted move- ment directions and are generally stable until they are undermined. we witl assume. Key blocks occur on the surface of an excavation and one or more of their faces are created by the excavation.4. orientation. permeability. but probably wave propagation. joint blocks without parallel faces are the rarity. The geological data that are required to determine the system of joint blocks are the spacing. It is rare that a key block will be formed with parallel faces. Not only comminution and excavation properties.4 Keyblocks (KB) and joint blocks (JB). for the time being. that particular values of spacing. joint blocks will usually have parallel faces. This distinction is more than surficial. blocks (JB) are rock blocks that do not contact the excavation surface. any rough- ness on the parallel surfaces creates rock bridges that must be broken before the block can move along them.3(b)]. On the other hand. and other system properties should depend at least partly on block shape. and orientation are assignable. In fact. and extent of each joint set. 5 or roughly cubical blocks [Fig. Although a mathematical discussion of block shapes may be ahead of engineer- ing practice. The joint Figure 5. The average spacing of joints in a set will dictate the average dimensions of the block per- pendicular to these joints. depending on relative joint spacings and orientations in the undisturbed rock mass. and statis- tical distributions are required to describe them. 5. as shown by blocks KB in Fig. grout take. For simplicity. extent. All rock mass properties that are affected by jointing are probably influ- enced by the shapes of the natural rock blocks in situ. These quantities are all nondeterministic. we expect that applications will emerge as the subject becomes established. The average extent of joints will dictate the probable sizes of the largest blocks.

Therefore. We now intro- duce the digit 2 to signify omission of a particular joint set.5(b)] signifies a block formed of joints 7. The latter requires the inter- section of a greater number of planes. blocks with fewer joint sets are more likely to occur than those involvin g alarger number of joint sets.2. it is labeled block 1100. block 1200 [Fig. First. such blocks are more stable than those lacking parallel faces because the directions .5(c)] is formed of joints 1. the joint pyramid is a spherical polygon. 5.5(a) depicts a block formed of four nonparallel joints. and 1200 will be more numerous than block 1100. Another intuitively derived principle says that a block with opposing faces formed by two joints of the same set is more likely to occur than a block formed without parallel joints of the same set. In general.3. the JP is an arc of a great circle. This block is the intersection of LlLrUr(In and. In the case of n nonparallel joints. With n . it is easier to accommodateo within an SP. and 4 only. and one joint set repeated to produce a block with one set of parallel faces. Intuition says that blocks 1120.5 Use of digit 2 to describe omission of a particular joint: (a) block 1100. and the probability thatn planes intersect each other varies inversely with r. Figure 5. 5.Chap. (c) block 1200. and 4 without any face from joint 3.1 nonparallel joints. an arc of a great circle than a spherical polygon. the block with parallel faces will tend to be more numerous in an excavation. so that block 1120 [Fig. using the digital notation introduced in Chapter 4. (b) block 1120. 5 Joint Blocks 129 sorts of joint blocks. (However. To create a finite block in an excavated space. Similarly. (a) (c) Figure 5. a JP must be contained in that space's SP.

5 of motion are greatly restricted. the joint-block system formed by a determined system ofjoints is relatively regular. and omitting joints 3 and a [Fig. they are probably more numerous inside the rock mass as well. JOIIIT BLOCKS IN TWO DIMETUSIONS Two-dimensional blocks build planar models of rock mass behavior. .6(a) can be transformed into block 3322by forming two faces from each of joints 1 and 2. blocks 3323 and 3322 are more likely than block 1120. We will begin with an analysis of the number of block types generated by z nonparallel joints. If this be so. that is. (c) block 3323.6(b)]. with no joints repeated in any block. yet for some a two-dimen- sional approximation proves sufficient. All real problems are three-dimensional. we would predict that 3322 is more likely than 3323.130 Joint Blocks Chap. a joint is a straight line in a plane P. We will use the digit 3 to indicate that a block is formed with both the upper and lower half-space of a given joint sef. which is more likely than 1120. Combining this with the previous principle. Joint Blocke When No Joints Are Repeated For analysis in two dimensions. To examine blocks bounded by a number of such joints we will appty the theory of Chapter 4 and judge the emptiness of the corresponding joint pyramids.2. (a) (b) (c) Figure 5. 5. In any case. and 4 and has no face ofjoint 3. 5. Similarly. According to our second intuitive principle. Thus block 1120 of Fig.6 Use of digit 3 to describe doubling of a particular joint: (a) block 1120. however. Some examples for discrete element analysis and finite element analysis appear in Chapter 1. the block shapes will not be prismatic unless the joints are mutually orthogonal. a two-dimensional discus- sion ofjoint blocks will serve as an informative introduction to the mathematical relationships required for three dimensions. (b) block 3322. block 3323 involves two each ofjoints 1.) If blocks with parallel faces are more numerous in an excavation.

we must move each to pass through the origin (0. the repeated joint is set 1. By the fi.7. the two finite blocks are the cousins 010 and 101.9. 5. We may regard these angles as two-dimensional analogs to Figure 5. the nonempty pyramids. without the stereographic projection. 5. they determine a pah of joint pyramids along the line of the repeated joint set as shown in Fig. the three joints define six nonempty joint pyramids. and therefore the infinite joint blocks are the three pairs of cousins: 000.2n joint blocks. Joint Blocks When One Set of Joints ls Repeated Now consider the possible joint blocks having n + I faces with only n joint sets. there must be 2" . so it includes both the upper and lower half-planes of the repeated joint. 111. with one joint set repeated. in two dimensions. and 001. There are 2" half- plane intersections with n joints. we can draw the joint pyramids directly.7.8. So. the finite blocks correspond to the empty joint pyramids. There are23 -2(3):2 flnite joint blocks. 100. the num- . When both of the latter are shifted to pass through the origin. 0) in plane P. shows three joint sets passing through (0. 0) that gen- erate six angles.011. respectively.7. According to the theorem of finiteness. 5. 5. that is. then n joints define 2n . 110.niteness theorem. Such a block is determined by the intersection of half-planes corresponding to all faces. When the number ofjoint sets forming a block is greater than 1. Since there are 2n joint pyramids. which uses the same three joint-set orientations as Fig. In the example of Fig. The n joints thus shifted will then create 2n angles subtended at the origin. for example. In this example. Continuing the ordered digital notation of Chapter 4.2n empty joint pyramids. Representatives of these block types are shown in Fig. By the process of elimination.7 Nonempty joint pyramids.Joint Blocks in Two Dimensions 131 Given n joints. joint pyramids. Figure 5. with 0 and I representing the half-planes above and below a joint.

which are the cousins 310 and 301.8.9 Joint pyramids with re- 1 peated joints. The remaining n . Again using the digit 3 to represent a repeated joint set. is 2. when n 2 2. . So there are only two nonempty joint pyramids. as in Fig. Suppose that one set out of n is repeated. 5. al --2 Figure 5. The number of empty joint pyramids is 22 .8 Finite cousins: (a) block 010. The number of empty joint pyramids is therefore 2n-t .10.2. When no set is repeated. Any one of these defines a nonempty joint pyramid if and only if it contains one or the other of the rays corresponding to the intersection of half-planes of the repeated joint set. So the number of nonempty joint pyramids. (b) block 101. 5 (a) Figure 5.9. Considering the intersection of all four half-planes. 5. there are only two nonempty joint pyramids.l sets define 2o-1 unique half-space intersections. 5. 2 : 2. three joint orientations uniquely determine the shapes of the finite blocks. It should be appreciated that the shapes of blocks with a repeated joint set now depend on the spacing of the repeated joint set. Typical blocks corresponding to the latter are shown in Fig. the cousins 300 and 311. let us examine the example of Fig. ber of nonempty intersections is not increased beyond the two rays identified by the repeated set because the condition that the repeated half-planes be inter- sected already limits the nonempty regions to these radii and they cannot be further subdivided by new radii from the origin. Joint Blocks Chap.

we pass all these joints through (0. 330. all of the nonempty joint pyramids must lie along joint 2 as well.11(a). The total number of joint-block types is then (2'-^)nUK" m)t. with m ) 2. If n:3 and m:3. we must conclude that there are nQ"-t .11(a)]. If set 1 is repeated. 5. For example. m!).0).11(b).10 Typical blocks belonging to the finite cousins: (a) 310. The number of half-space intersections produced by the nonrepeated joints is 2o-*. all of the nonempty joint pyramids must lie in its line.2 joint blocks. This is true for any additional repeated joint sets.times the number of combinations of n joints taken m at a time (Cf). (b) 301. Consider now the different types of joint blocks that could be fashioned from n joint sets when only one joint set is repeated. 5. Therefore. Continuing in this fashion through all possibilities. There are then (23-2)3UQl t!) : 6 types ofjoint blocks corresponding to 033.0). as shown in Fig. 2) individual joint-block types. the number of unique joint-block types when m joints out of n are repeated is 2"-. Therefore. and 331.Joint Blocks in Two Dimensions 133 Figure 5. But if the second set is repeated.2 joint blocks. suppose that - n : 3 and m: 2. Joint Blocks When Two or More Sets of Joints Are Repeated Let there be n sets of joints of which m are repeated. This many joint blocks are created by assigning the first z joints as repeated and an equal number is created by selecting any other combinations of z joints as repeated. To construct joint pyramids. and therefore in the intersection point of the two repeated joints [Fig. there are correspondingly 2o(3UQ! 3!)): l. all the possible joint pyramids are empty. as before. So the number of finite joint blocks is 2"-*. 133. . Given that the first joint set is repeated yields a corresponding 2"-t . 5. namely block 333. which is always point (0. Proceeding to make the second joint set the only repeated one yields another 2"-t . 303. which is shown in Fig. 313.

block 3L20 has cousin 3021. If a joint block is finite.1 are even numbers. An incremental analysis is used in which we determine the number of half-spaces added to the existing number when an additional joint is allowed to cut through the origin. repeated joint sets. JOI]UT BLOCKS IN THREE DIMETIISIONS The types of joint blocks in the three-dimensional world will now be discussed formally. For example. . These are sum- marized in Table 5. We will accumulate the number of nonempty joint pyramids formed by a series of joints as they are added one by one to the list. and vice versa.1. 5 (b) (a) Figure 5. It will be noted that all the entries of Table 5. Formulas for Joint Blocks in Two Dimensions Expressions for the numbers of joint-block types in two dimensions have now been developed for all possible cases of repeated joints. its cousin is finite. the joint pyramid is a solid angle made by a series of planes passing through the origin (0. The interchange of 0 and 1 to form cousins does not alter 3's and 2's. First we consider blocks having no repeated joints as faces and then extend these results to cases with one or more repeated joint sets. This is because all block types come in pairs since the theorem of symmetry assures that every block has a cousin of the same type formed by changing its 0's into 1's.0). the results of the preceding section can be incorporated. Joint Blocks When No Joints Are Repeated In three dimensions. 0. In this wily.1U Joint Blocks Chap.11 Multiple.

-.. n ml(n-m)! 1.1 Numbers of Block Types in Two Dimensions Number of Empty Number of All Combinations Number of Non.2) n) 2 rn definite repeated sets 2n-^ 20-.| 2n n(z-r .'.:@.) ol . TABLE 5.m (. Joint Pyramids of Half-Planes (All Joint empty Joint (Number of Finite Number of Repeated Sets Pyramids) Pyramids Joint Blocks) Condition 0 repeated sets 2" 2n 2 -2n n)l I definite repeated set 2.2.)n-m nlm)2 -n (m2D Qf choices) Note:C!: rrnl .7n-m cm.| 2 zn-r. 2 n)2 Any I repeated (n choices) set nTo. nlm)Z (m> 2) Any m repeated sets f m.

The total - number of nonempty joint pyramids is then .12 Nonempty joint pyramids with up to three planes.. n joints (n > l).13 Nonempty joint pyramids with addition of a fourth prane.1). Addition of a second joint to the first subdivides each of the previous joint pyramids. The previous joints cut plane 3 into four angles so that the addition of plane 3 must provide a total of four additional joint pyramids. The nth joint plane is cut into 2(n 1) angles by the previous (n - 1) joints. Two joints. Three joints: Addition of a third joinf to the first two subdivides the pyramids as shown in Fig. 5. (a) (b) Figure 5. There are now 8. . .13).136 Joint Blocks Chap.12(a)]. the addition of the nth joint to the - first (n . or 2 * 2(2 . This may be written as 2(3 . Therefore. - Plane 1 I t/ I \l I tl L__ __ 7_:_---*---) -_J_ Plane 2 Figure 5. nonempty joint pyramids (Fig.l) + 2(3 1). 5 one joint: A single joint divides the whole space into two joint pyramids. and the total number of joint pyramids is 4.12(b). Thus there are two additional nonempty joint pyramids [Figure 5. 5.1) joints yields 2(n 1) additional joint pyramids.

the number of nonempty joint pyramids is 2(n . the total number of empty joint pyramids is 2h-') 2(n 1). with n ) 2. Two repeated joint sets.This is then the number of joint block types with joint sets and n joint-block faces. Therefore.n t 2). Assume that the first joint set is repeated.L/ n 2(i-r):2[r + i=1 Er] n(n.15). the number of empty joint pyramids is 2n . -\. there are two nonempty joint pyramids. 5.1) angles (as noted in the preceding section).14 Nonempty joint pyramids with one repeated joint set.Joint Blocks in Three Dimensions n-l ?_L\- | .14). Each nonempty joint pyramid is therefore an angle subtended at the origin and measured in plane I between two limiting joint planes of the other sets (Fig.' -n-Lz The number of combinations of half-spaces is 2". Therefore. 5. Since there are nz . Therefore.1 planes 2 through n will cut plane 1 into 2(n . any nonempty joint pyramid must lie simultaneously in the planes of joint set 1 and joint set 2 and therefore along their line of intersection. Planes 1 and 2 are repeated but planes 3 to n . n Joint Blocks with n Sets of Joints and One or More Repeated Joint Sets One repeated joint set. .l)1. which passes through the origin.n f 2 nonempty joint pyramids. Figure 5.2(n . - Considering each set in turn as the repeated joint set. The total combinations of the (n . divides this intersection line into two rays from the origin (Fig.(n' . Plane 3. the n .l)\:r. If there are more than two joint sets (n ) 3) and both sets 1 and 2 are repeated. there will be n times this many types of empty joint pyramids.l). The total number of types of finite joint blocks is therefore equal to nl2r'-t) .1) joint planes of sets 2to n is 2(n-1).t :z(1 2 / '. Then any nonempty joint pyramid is simultaneously in the upper and lower half-space of joint plane 1. If there are njoint sets.

This is the number of empty joint pyramids. The number of combinations of the upper and lower half-spaces of the (m + 1) to n nonrepeated joints is )tn-n). Considering all possible combinations of n joints of which 2 are repeated gives as the total number of empty joint pyramids c!1}r"-zt :n(n:2 l)r2h-2.4) shown in Table 5.)("-m). the number of empty joint pyramids.2. More than two repeated joint sets.2 summarizes the numbers of joint pyramids for the qases con- sidered. STEREOGRAPHIC PROJECTIOIU SOTUTION FOR JOITIT BLOCKS Using the formulas in Table 5. we are now in a position to explore a three- dimensional example. the total number of combinations of half-spaces in this case is 2b-D. . Considering all possible choices of m repeated sets from n joint sets yields the total number of joint blocks as Ci.15 Nonempty joint pyramids with two repeated joint sets.2. each nonempty joint pyramid must contain the line of intersections of the repeated joint sets. with m) 3. are not. In the following section we use this table to identify all the joint pyramids on the stereographic projection. These are based on the sy$tem of four joint sets (n . 5 Line of intersection ofl and2 Figure 5. 2l . and therefore the number of joint blocks when the first m joint sets are repeated. We will discuss all the joint pyramids in every case of Table 5. If the number of repeated sets is m. n(n .3. 1]. Therefore. 1)[2<'-st . This was the result upon choosing joints 1 and 2 as the repeated sets. Joint Blocks Chap. is 2h-2) .2. corresponding to a selected example. Summary of Formulas for Joint Blocks in Three Dimensions Table 5. Accordingly. and therefore the number of finite joint blocks. This means that only the origin can be contained in such joint pyramids.2) planes. 21 - or finally. Since we can choose either half-space of the latter (n . all joint pyramids are empty in this case.

2@ . TABLE 5.lX2"-3 .2@ . Number of Joint of Half-Spaces (All Joint empty Joint Pyramids (Joint Number of Repeated Sets Pyramids) Pyramids Blocks) Condition 0 repeated set 2n n2-n*2 2-(rf -n*2) n>l I selected repeated set 2r-r 2(n .t) n)Z Any I repeated set n2-r Zn(n .l) n(n .2 Number of All Combinations Number of Non.l)2-t n(n .l\ zo-t .l)l n) Z 2 selected repeated sets 2-z 7 2-z -2 n)j Any 2 repeated sets n(n ."-3 sets gt (D .l) nf2n-1 .1) n) 3 m selected repeated sets 2n-^ 0 2n-^ n) mi_3 (rn > 3) Any m (m> 3) repeated ?m)n-m 0 C!2-^ n) m .

5. Label each segment of each circle.3 Dip. p Set (des) (deg) I 75 80 2 65 330 3 40 30 4 10 270 Finite Joint Blocks with No Repeated Joint Sets 1. the upper half-space of set 2. For example. region 1010 corresponds to the intersection of the lower half-space of set 1. Since we have used a lower-focal-point projection. region 1010 is the spherical surface outside of circle 1. 3.the lower half-space of set 3. This has been done in Fig.16(a). Considering the binary list developed above. 2. a Dip Direction. Finite Joint Blocks with One Repeated Joint Set 1.L4: 2.3.2 as nz . and inside circle 4. 5. Identify in order all the combinations of upper and lower half-spaces of the four joint planes and indicate by the letters JP the fact that a particular joint pyramid has been located on the stereographic projection. This yields the following binary list: 0000 JP 0110 JP 1100 JP 0001 JP 0111 JP 1101 0010 1000 JP 1110 JP - - 0011 JP 1001 JP 1111 JP 0100 JP 1010 JP 0101 JP 1011 JP The number of combinations of half-spaces for the four joints is 2a : 16. inside circle 2. 1030 identifies the cir- cular arc segment along circle 3 that is outside circle 1. inside circle 2. Blocks corresponding to these are shown for equal spacing between joints of each set in Fig. the empty joint pyramids are 0010 and 1101.n + 2: 14 and the number of empty joint pyramids is 16 . The number of joint pyramids (JP) is given by Table 5. They are shown in Fig.16(b) and (c). . The digit 3 in the place ofjoint i means that the segment belongs to circle i. 5. outside circle 3. For example. Joint Blocks Chap. Compute the radius and center of each projection circle and draw the four circles for Table 5. First compute the radius and center of projection of the circles of each of the four sets. and inside circle 4. 2. and the upper half- space of set 4.17(a). Number each region delimited by circles. 5 TABLE 5.

. (b) block corresponding to empty joint pyramid 0010. Reference circle Figure 5.16 Analysis of joint blocks with no repeated sets: (a) joint pyramids with no repeated joint set. (c) block corresponding to empty JP ll0l.

.1 4 --.17 Analysis of joint blocks with one repeated set: (a) JPs with one repeated joint set. (c) block corre- sponding to empty JP 3101. Joint Blocks Chap. \\ (b) Figure 5. (b) block corresponding to empty JP 3010. 5 -:.

18(b) and (c).2 is ){n-1) :23 :8. assume that set I is repeated. From this table. Therefore. For example. Finite Blocks with Two Repeated Joint Sets 1. . The empty joint pyramids can be identified from the list as 3010 and 3101. These are drawn assuming all joints have equal spacing. A tally of this procedure is given in Table 5.4 and identifies the empty joint pyramids as 0013 and 1103. These are nonempty joint pyramids. and therefore finite joint blocks.3. so there arc 12 joint block types with two faces repeated. 0331 means the intersection point of circle 2 and circle 3.1) : 6. 5. determining the empty joint pyramids as 0030 and 1131. this direction is inside circle 2 and outside circle 4. 2. All of these empty joint pyramids represent finite joint blocks. Now identify the joint pyramid number corresponding to each intersection point. the number of nonempty joint pyramids with set 1 selected as repeated is 2(n . The list of these eight combinations is as follows: 3000 JP 3100 JP 3001 JP 3101 3010 3110 - JP 3011 JP . Following the same procedure. inspection of Fig. Similarly. 3. the number of empty joint pyramids is 8 . a choice of joint 4 as the repeated set produces the list of joint pyramids in the fifth column of Table 5.16(a) yields the list of half-space combinations shown in the third column of Table 5. 5. Blocks corresponding to these have been drawn for the condition of equal joint spacing in each set in Fig. For example.5. 3111 JP According to Table 5. the empty joint pyramids are identified as 0310 and 1301. All the others are empty joint pyramids.4. The two finite blocks 3301 and 3310 for the case of joint sets 1 and 2 repeated are shown in Fig. Pursuing a similar analysis with joint set 3 as the only repeated set produces the fourth column of Table 5. There are two empty joint pyramids for each specific choice of repeated joint sets and six such cases (C? : 6).6: 2. Finally. which is inside circle L and outside circle 4.2. Identify in order all the combinations of upper and lower half-spaces with a given repeated joint set and indicate by the letters JP the fact that a particular segment has been located on the stereographic projection.17(b) and (c). The preceding analysis was based on the repeated joint being from set L.4. Again prepare a stereographic projection for the four joint sets of Table 5. Now assume that set 2 is repeated. Identify in order all combinations of half-spaces with two planes repeated and note which ones appear as intersection points on the stereographic projection. The number of combinations of half-spaces stated in Table 5. 3031 identifies the intersection of circles 1 and 3. 5.Stereographic Projection Solution for Joint Blocks 143 3.

3r No Joints Repeated Joint Set I Repeated Joint Set 2 Repeated Joint Set 3 Repeated Joint Set 4 Repeated 0000 3000 0300 0030 empty 0003 0001 3001 0301 0031 0013 empty 0010 empty 3010 empty 0310 empty 0130 0103 0011 3011 03r1 0131 0113 0100 3100 1300 1030 1003 0101 3101 empty 1301 empty 1031 t0t3 0110 3110 l3l0 ll30 1103 empty 01ll 3111 l3l I l13l empty 1113 1000 1001 1010 1011 il00 1101 empty 1110 1111 rAIl the empty joint pyramids are labeled "empty.4 Empty and Nonempty Joint Pyramids with No or One Repeated Joint Sets for the Joint System of Table 5. .D 5 TABLE 5." The other joint pyramids can be found on the stereographic projection.

I . (b) block corresponding to empty JP 3301.. (c) block corre- sponding to empty JP 3310.18 Analysis of joint blocks with two repeated sets: (a) JPs with two repeated joint sets. 1 3 Figure 5.Stereographic Projection Solution for Joint Blocks 145 tt.

3331 Figure Empty and Nonempty Joint Pyramids with Two Repeated Joint Sets for the Joint System of Table 5. the only half-space intersection is 3333.4 0333.3133 1.3. . If all four joint sets are repeated (* :4). 5 TABLE 5.20 shows the finite block 3333.6 lists all such combinations.2.6 Empty Joint Pyramids with Three Repeated Joint Sets for the Joint System of Table 5. (b) 3331.19 shows the finite blocks 3330 and 331.4 3303. This combination forms an empty joint pyramid. Figure 5.1333 1.19 Finite blocks corresponding to: (a) 3330. Joint Blocks Ghap.3 Repeated Joint Sets Empty Joint Pyramids 2. Table 5.4 3033. Figure 5. TABLE 5.3 Repeated Joint Sets land2 land3 land4 2and3 2and4 3and4 3300 3030 empty 3003 0330 empty 0303 0033 empty 3301 empty 3031 3013 empty 0331 0313 empty 0133 3310 3130 empty 3103 empty 1330 1303 empty 1033 3311 empty 3l 31 3113 1331 empty 1313 1133 empty Finito Joint Blocks with Three or More Repeated Joint Sets All combinations of half-spaces with three or more joint sets repeated form empty joint pyramids.3 3330.3313 1.

so are 3110 and 3001. 5. related to each other by a center of symmetry.95r2 o. Thejoint blocks corresponding to these joint pyramids are cousins. --- \\. 0030 and 1131. Symmetry of Joint Pyramids In all cases all the combinations of half-space intersections can be divided into symmetric pairs-for example: 0100 and 1011. and 3303 and 3313. and 5. -.7848 o.76ffi 4 -0. Since all combinations of joint half-spaces can be divided into symmetric pairs. TABLE 5.3213 0.7 Joint Set I 0.1 COMPUTATION OF EMPTINESS OF JOINT PYFAMIDS USIIIG VEGTORS Consider the four sets of joints of Table 5.1736 0 0.16(b) and (c). For example. The corresponding upward normal vectors are shown in Table 5.\_ '-_ \\ \.19(a) and (b).--4 1 \.t677 0. the empty joint pyramids must also possess the same symmetry. If a combination of digits defines a nonempty joint pyramid.4226 3 0.20 Finite block 3333.Computation of Emptiness of Joint Pyramids Using Vectors tr. and 0303 and 1313.5566 o.7.3.0313 and 1303. [Compare: Figs.2588 2 -0.4531 0. For example. and 0133 and 1033.18(b) and (c). 5. 5.l-- -_ Figure 5. 0010 and 1101 are both empty joint pyramids. So are 0310 and 1301.9848 . 0000 and 111L are nonempty.17(b) and (c). changing the number of the joint pyramid by interchanging 0's and 1's produces another nonempty joint pyramid.

-Z) satisfies all of inequalities (5.6398 -o.42262 (5.1347 -0. Y.1677 v + 0. The edges of joint pyramid 0100 are Lines of Intersection Sign of (X.3 214 i + 3. the joint pyramid is empty.Z) and (.Y.4641 214 0.s566 Y . For each line of Table 5.98482> 0 We will compute the edges of joint pyramid 0100.6125 0. .8378 113 0.1677Y + 0. 0.1) 0.7848 Y + 0. therefore.95t2X+ 0. 0100 is an empty joint pyramid.9st2x + 0.r736X + or+ 4.76602 > 0 -0. Y. (X.8.X. .8.7848 Y . The coordinate inequalities of 0100 are 0. Those vectors that satisfy inequalities (5.1).0194 0. -Z) into inequalities (5. no vector (X.Y.3213X + 0.X.2s882>0 -0. Y.8049 -0.2 -o.>00.0J6602 > 0 -0.8895 0. -Z) are two TABLE 5.1).e857 0.1658 -0.4291 0.25882 >0 0.1568 314 -0.X.8 Lines of Intersection of the Joint Planes Intersected Planes 1.4 If there is no vectorsatisfying (5.1736X + 0r+ 0.5930 lr4 0.4s3rx . Z) and (.2 1.6282 -0. Y. .1) are the edges of the joint pyramid 0100. .0292 2r3 --0.2) -0.4531X + 0.5289 0.3213 X - 0.42262 > 0 (s. First all the intersection lines of these four planes are calculated and listed in Table 5. Substitute each (X. Z) 1.2).98482> o For each line of Table 1. Y.7661 0.1350 opposite vectors in the corresponding intersection line. its inequalities are 0.148 Joint Blocks Chap.5566 Y + 0. Z) or (. Consider joint pyramid 0010. 5 Computation of Joint Pyramids with No Repeated Joint Sets Consider joint pyramid 0100.

r736X + OY . 0. Substitute (X.7 we need to consider only the intersection lines 12.8049.0. These two vectors are therefore the edges of joint pyramid 1310. -0. and in Table 5.0194.5930) Using the same method.8049. Using the same method.Y.5566 y _ 0. We do not need to compute joint pyramids with three or more repeated joint sets since all such joint pyramids are empty.95l2X+ 0. 0. 0. Y.98482> 0 Since joint pyramid 3031 is a ray of intersection line 13.4531X + 0.98482> 0 Joint pyramid 1310 is a sector of plane 2. Z) and (. .42262 > 0 (5'4) o.9512X .25882 : O -0. -Y.0194. -0. Z) (0.32r3X- -0. -Z) of 24 satisfy equations (5.5930) The second vector satisfies (5. The coordinate equations of 1310 are -0.0.42262 : O (5'3) 0. Y.23. In this section we demonstrate the steps in a hypothetical calculation and suggest potential applications relevant to energy requirements for rock comminution. we will find that joint pyramid 3030 is empty.7848 Y+ 0. and 24.X. Therefore.4s3rx + 0. joint-block theory is new and there are no examples of its actual application.1: (X.3). 0.1677 r + 0. APPLICATIONS OF BTOCK THEOBY: AN EXAMPLE As noted in the introduction to this chapter. .Applications of Block Theory: An Example 149 Gomputation of Joint Pyramids with One Repeated Joint Set Consider the joint pyramid 1310. and 24 into equations (5.23. Vectors (X.32r3x+0.5566 Y+0-76602:o 0.Z) of 23 and (-X.5930) (-X. we will find that joint pyramid 3010 is empty. :Y. we only need to test the two opposite vectors of intersection line 13 of Table 5. so joint pyramid 3031 is the ray determined by vector (-0.0194. -Z) of intersection lines 12.-Y. The coordinate inequalities of 3031 are O.3). -Z): (-0. -0.8049.76602>O -0. Gomputation of a Joint Pyramid with Two Repeated Joint Sets Consider the joint pyramid 3031.1).L677 r.7848 Y + 0.25882 > 0 -0. the edges of 1301 are intersection vectors of plane 2 with other planes.r736X + 0r+ 0.

21). The true joint spacing is measured in the direction of the normal to the joint set. there will be blocks in the rock mass corresponding to all the empty joint pyramids. a reference line is laid across an enlarged photograph of an outcrop. If each joint set has close spacing and long continuation (extent).4). Joint Spacing A statistical evaluation procedure for joint spacings was presented in papers by Priest and Hudson (1976. it is possible to make a formal analysis of a system of n joint sets to characterize the spacing distribution parameters of every set. sinarcos(8. however.2I. 5. 1981) and Hudson and Priest (1979) for which data were obtained by "scan-line surveys. One has to know the dip and dip direction of both the joint set and the plane of observation. one then measures the frequency at which joints cross the reference line. sin a1 cos p. the line of intersection of the two planes. Whether or not a scan-line survey is made. d Jl sin cos2 d - Therefore. In such cases.. from which d:J-=:L -. is 6.3 with four joint sets (n . More often." For a scan-line survey. . It is possible to calculate the true joint spacing from the apparent spacing measured on a rock face.. a section perpendicular to the line of intersection of a joint plane (1) and a rock face (2). joint spacing values can be determined directly from outcrops or excavations. one or more joint sets has limited extent and all the sets are spaced differently. Priest and Hudson concluded that most joint-frequency data are fit by the negative expo- nential distribution. cos p2.150 Joint Blocks Chap. some of the empty joint pyramids will lack corresponding joint blocks. cos ar) where & are B the dip and dip direction of the respective planes. while at least one of the other sets has wide spacing and limited extent. 5. The procedure for calculating the true spacing from the apparent spacing and planar orienta- tions is based on Fig. . about n. The normal vectors of planes 1 and 2 are Dr : (sin a. Often one set. and obtaining data from at least n nonparallel scan lines or drill holes.2. Using this distribution as a model. The true spacing can therefore be seen by rotating f . calculated as follows: cosd: nl . The angle between n. sin c. (Fig. 5 For the example problem we will reuse the data of Table 5. The apparent spacing is the separation of parallel joint traces in a plane that does not contain the normal to the joint planes. n2: sina. has close spacing and long extent.f") * cosdl cosd2 The apparent spacing d is measured normal to I. The true joint spacing is ft and the apparent spacing of plane 1 measured in plane 2 is d. sin pr. cos d1) [2 : (sin a. sin fr. related to bedding or foliation. and n.

Length in u fl in Plane 5 h Plane 5 Plane (dee) (dee) (m) (m) (m) I 75 80 2.6) Equation (5. plane I Figure 5.5) 4 -Itloai sind2cos(Br or h:d (5.5 3 Q 30 5.5 4 10 270 3.8 5 60 50 .38 6.33 9.8 t.Applications of Block Theory: An Example 151 Joint.* cos il" ^'=. Spacing (d) Spacing.46 Very long ) 65 330 l0 9. We will illustrate both procedures in a hypo- thetical example. Equation (5. Table 5.24 7.9 lists the dips and dip directions of the four sets of joints previ- TABLE 5. Direction.6) can be used to calculate the true spacing from joint traces seen in a rock face of known orientation.9 Dip Apparent Calculated Trace Dip.5) can be used to calculate the separation of joint traces in a given section through a rock mass when the joint spacing is already known. d: n| =* .5 3.21 Section perpendicular to the intersection of a joint plane and a rock face.5 2.zlz (5.

were calculated from equation (5. 5 ously analyzed. is to plot the stereographic projection. we can also calculate the apparent spacing values in any other section plane. together with the dip and dip direction of a section plane. or the roof or wall of an actual underground gallery. . the wall. for a generally inclined section plane. Gonstructing a Trace Map in Any Section Plane The first step in constructing a trace ffi&p. 5. denoted as plane 5. I 4 1000 /I a\ i.2| Lower-focal-point stereographic projection of data in Table 5. From these spacings. This has been done in Fig. The true spacing values (ft)./ 1 rdl Reference circle Figure 5.1 \-d' 1ro tor Otll .9. or an ex- ploratory excavation. On plane 5 we proceed to measure the spacings and trace lengths of the joints of each set. we can now draw the various joint blocks generated with different numbers of repeated joint faces.162 Joint Blocks Chap. Plane 5 may represent a rock outcrop. Knowing the true spacing values of the different joint sets.b \ I \.22. t\ \ \ .6) usinE crz : 60o and f z : 50". corresponding to the apparent spacings and orientation data.

* Knowing the apparent spacings and the half-spaces of all the joint traces. column 3). we can now draw a section in plane 5 and locate all the finite joint blocks. It may be considered a generic trace map. 5. 5. it will not be obvious which side of a joint trace corresponds to the upper half-plane (U) and which to the lower (I) when seen in the section. Obtaining such data is not as simple as obtaining joint-spacing data. making use of the added information on joint extent hypothesized in Table 5. and Warburton (1980). In like manner we find that the lower side of joint traces 2. and 4 all correspond to the lower half-spaces of these planes. In order to discuss the joint blocks cut by the section. To determined the name of a joint pyramid.23 is only one possible solution from an. In Table 5. the lower side of trace 1 corresponds to the upper half-space of plane 1. 5. A section has been drawn in Fig. Next we must determine the apparent dip of each plane in the section and draw the trace orientations in the plane of the section (Fig. Plane L corresponds to bedding and therefore set 1 is assumed to have almost infinite extent. Since there is freedom in placing the joints of finite extent within the sec- tion. and the reference circle. They therefore represent the two opposite horizontal vectors in the plane of the section. 5. this is readily established by the following proce- dure.22 d is the projection of vector 4 the dip vector of plane 5. infinite population. On Fig.10 Half-Space Corre- Location of Point d Half-Space sponding to the Lower Plane with Reference to the Corresponding Side of the Joint (Great Circle) Great Circle tod Trace in the Section I Inside U U 2 Outside L L 3 Outside L L 4 Outside L L In the section plane. we made use of the labeling of half-spaces carried out previ- *An alternative method of identifying (/ and. However. we have assigned a number to every closed polygon of Fig.23). as discussed by Priest and Hudson (1981). the direction corresponding to d is now identified as either L or U (Table 5.23. In Fig. Therefore.23.10. and there is a statistical problem with bias and truncation. The other joints have finite extent.3. plane (5). Baecher and Lanney (1978).Applications of Block Theory: An Example 153 The section plane (5) being inclined.9.22. The direction of / was determined to be in the upper half- space of plane 1.L in the trace map is presented in the last section of Chapter 6. . we determine in which half-space of each joint plane the point d lies. TABLE 5.10. The results of this step are given in Table 5. Fig. points a and b mark the intersections of the section.11 the joint pyramids corresponding to each of these polygons are named. 5. 5.

..(n $ 72. looking toward the southwest.1" ).23 Section in plane 5. \ 68i1" r\ Figure 5.

Applications of Block Theory: An Example 155 TABLE 5. has been labeled 3210. 5. polygon 20. created by a pair of planes of set 1. Thus polygon 4.23) Polygon Number Joint Probable Joint Finite or on Fig. every joint pyramid should have its first digit replaced by 3. meaning that every . 5. This is certainly expectable in the case of bedding (joint set 1). Therefore. behind the section. involving only pairs of traces of planes 1 and 4.23 Pyramid Block Infinite? I 0201 3201 Infinite 2 I l0l 3t0l Finite 3 lt23 3103 Finite 4 3Z2l 3223 Infinite 5 1t0l 3r01 Finite 6 0010 30r 3 Finite 7 0023 3003 Infinite 8 0001 3001 Infinite 9 3212 321 1 Infinite l0 3223 3r23 Infinite ll 3201 3001 Infinite l2 32lO 3013 Finite l3 3223 3023 Infinite l4 tt23 3123 lnfinite l5 3223 3203 Infinite l6 3223 3123 Infinite t7 3223 3213 Infinite l8 oo23 3013 Finite l9 3201 3001 Infinite 20 3210 32t3 Infinite 2l I I0l 3103 Finite 22 0003 3003 Infinite 23 rzlo 3l 13 Infinite 24 0201 3001 Infinite 25 3210 3210 Infinite 26 ll0l 3103 Finite 27 0001 3003 Infinite 28 t 1l0 3113 Infinite 29 0010 3013 Finite 30 3201 3201 Infinite 31 3223 3223 Infinite 32 3ZO3 3003 Infinite 33 1210 3110 Infinite 34 3201 3ZO3 Infinite 35 3210 3213 Infinite 36 fi23 3l l3 Infinite 37 0023 301 3 Finite ously. Up to now we have ignored the possibility that a joint whose trace is out- side the polygon might truncate the underside of a joint block.f l Joint Pyramids and Joint Blocks fntersected bY the Section (Fig. is labeied as 3223. the lower side of trace 3 and the upper side of trace 4.

11. However. assisted by drawings of the different joint blocks as seen looking at the section. . by methodical study of all the polygons. Recognition of the likelihood that bedding will qeatetwo faces of this block requires changing the label to 3023. This rock mass thenhas rela- tively few finite blocks and the rock mass is semicontinuous in nature. the consumption of explosives per unit volume of rock) should be rather large.11 the section parallel to plane 5 intersection 37 complete polygons of which 11 were finite and26 infinite. For this analysis.. nonintersecting traces will suggest other modifications to recognize that some other joint will slice that block behind the section. From the point of rock communution. Polygon 7 was labeled initially as 0023. the powder factor (i. with completely finite joint blocks inside the rock. The Ratio of Finite to Infinite Blocks as a Rock Mass Index From Table 5. Digit 2 means that joint plane 3 does not bound polygon 7 in the section.4 and 5. the joint blocks corresponding to the joint pyramids are found to be those listed in the third column of Table 5. Comparison of the joint-block numbers with previous results of this chapter (Tables 5.e. Progres- sive collapse of tunnels or slopes following loss of a key block is less likely than in a fully discontinuous rock mass. Thus the proper joint block is not 3023 but 3003.156 Joint Blocks Chap. 5 joint block lies between two bedding planes. In this w&y. the trace map shows that a joint plane of set 3 lies just below polygon 7. Close inspection of every polygon with respect to proximate. it is necessary to apply judgment.5) determines whether the block is finite or infinite.

but the influences of added fractures in by the u-. parallel to their line of intersection. 167 . Here we concentrate on be discussed in Chapter 8 in connection of blockr into excavations that are entirely at the ground surface' movement Failure Modes Rock slopes present an infinity of failure modes' But on close analysis' these derive from mixtures of a few fundamental modes : sliding of a block along one face .1(e). and for abutments of dams [Fig. railroads. and roads tFig' 6'l(b)l' Excavations 6. canals.1(d)]. chapter 6 Block Thoory for Surficial Excavations BASIG CONCEPTS for In this chapter we explore the application of block theory to engineering from small rock surficial excavations. sliding of a block along two faces. Cuts into rock. stability of rock masses adjacent to neering designer must address not only the forces and the effects imposed a sloping free surface. factories.tion of the structure and the percolation of water through the rock. range concrete structures' Rock faces to those whose value rivals that of impressive and powerhouses [Fig' 6'1(a)] excavations provide space for buildings.1(c)J' for spill- on the sides of hills are made for foundations of bridges [Fig. The portal design problem will with tunnels. 6. In the latter works. Rock excavations are also made to gain access to undergtound as in Fig. 6. for various purposes. and for routes of pipeiines. the engi- ways. Stability problems are common in portals to the underground create addi- because the intersections of underground and surface excavations tional freedoms for movement of iock blocks. openings.

(c) as a bridge foundation. These may slip in progression. 158 Block Theory for Surficial Excavations Chap. and fracture ofthe rock because ofshear or bending stresses. or simultane- ously in a complexly shaped body.l Rock slopes: (a) for a powerhouse. 6 Tunnel (e) Figure6. In many rock masses.2(a) shows a typical wedge failure in which a single block slides along two joint planes simultaneously. the surfaces of sliding create multiple. advancing in the direction of the line of intersection of the two surfaces. rotation ofa block about an edge. The direction of movement must be parallel to the line of intersection of the sliding surfaces. (d) as a dam abutment. Figure 6. parallel lines of intersection. because this is the only direction in space that is shared by both surfaces. (b) for a highway. Figure 6. (e) as the portal of a tunnel. forming edges of a series of similar wedges. .2(b) shows a rock slope after the removal of a series of geometrically similar wedges.

or curved as in Fig.2 Wedge sliding: (a) a single (b) a slope after removal of a united wedge.3(b) and (c).Basic Concepts \\ \ \ Figure 6. 6. the actual block in motion may glide along different planes with that orientation. Although only one orientation of sliding surface is involved. Another fundamental failure mode for rock slopes is block rotation. An . either planar.3(d). 6.3(a). 6. as in Fig. A second fundamental mode of failure is sliding along a single rock face. as in Fig.

(c) on a con@ve fa@. 160 .3 Single-face sliding: (a) on a planar face. (c) Figure6. (b) on a convex face.

How- ever. ld) example of a slope failure involving mainly block rotation is "toppling failure. this arises naturally under gravity alone because the layers incline into the hillside. Failure of slender columns of rock occurs because each column receives an overturning moment about its base.4.4.Basic Concepts 161 Figure 6. .4 Sliding and toppling of multiple blocks. 6. 6. columns with other orientations can overturn in response to ice or water Figure 6.3 (Continued) (d) A united block." depicted in Fig. In Fig.

It cannot be denied that some failures could not have been prevented by a simple key-block approach. In this case it is quite likely that all movement would have been curtailed had block I been restrained. however. and can be prevented by retention of a key block. G forces. or forces transmitted from adjacent rock masses.. 6. [n columns lacking cross joints. the bending of the toppling columns was facilitated by opening of previously existing cross joints. Key-block Analysis Potential key blocks of an excavation differ in important respects from finite joint blocks within the body of a rock mass.162 Block Theory for Surficial Excavations Chap. In this work. such as (l) of Fig. For example.6 has failed by coupled movements of four blocks. 6.5 shows a block that has cracked in bending after initial sliding along an edge. when modes nessitatfig new fracturing or shearing through intact material are invalidated as candidates for key-block analysis. Figure 6.5 Beam bending. This may occur by virtue of the stress concentrations at the boundaries of the excavation. we assume that there is a set of key blocks and that more complex failure modes that may progress from their movement need not be considered as long as the key block is kept in place. Complex failure modes like this tend to develop progressively. Actual rock slope failures often encompass combinations of the funda- mental modes and may involve some new fracturing as well. a . The list of exceptions is shortened. It will be the reader's responsibility to judge the validity of this assumption in any practical case. For example. 6. Excavations at the surface also create mechanisms for new fractures to grow in rock masses. the slope in Fig. the bending leads to flexural cracks. each with distinct mechanisms.e. or in response to bending stresses accom- panying movements of rock blocks.7.4. Figure 6. toppling of steeply dipping layers is often induced by downslope sliding of soil layers across the bedrock surface. possess at least one face belonging to the excavation surface (i. Key blocks. In Fig.

(3) joint block. (2) removable block with parallel faces. . Figure 6.6 Progressive slope failure.7 Types of blocks in a slope: (l) key block.Basic Concepts 163 6 Figure 6.

. the methods to be developed in this chapter are different from those of Chapter 5. In this procedure. At this step ofoutput. is intersected with the other half-spaces (i. do not. and L. 6. for potential key blocks. the spacings might be described by an appropriate distribution (e..7. At the initial stage. such as (2) of Fig.g. the key blocks will include some with nonconvex shapes formed by the union of convex key blocks.e. (1983). If a joint block is cut through by an excavation surface. or a removable block having one or more pairs of parallel faces. only the mean orientation. subsequent analysis can identify the most critical key blocks. nor L. The . the key-block analysis can then determine the actuallocations of key blocks on the surface. A distribution of joint extent could also be introduced to calculate more accurately the probability of any key-block type having a volume larger than a given size. or 3. 2. would be entered. The spacing of eachjointset is input ind the volumes. the joint-set orientations might be specified in terms of parameters of a spatial distribution of planes about a mean joint-set orientation. whereas joint blocks. Analysis of key blocks in the generated trace maps. For this reason. each block is identified by a code (Dr) comprised of a string of numbers digit of D" is 2 if neither (1. can be used to calculate the dimensions and support . Finally. it tends to produce either a tapered block that is not removable. simulation procedures can be used to generate hypothetical trace maps on the excavation surface. The analysis for locating key blocks of a surface excavation is initially geometric. 6 "free surface"). together with Monte Carlo repetition. After identifying the block codes D. A program for statistical simulation of trace maps and subsequent key-block analysis was reported by Chan and Goodman.e. the first input being the dips and dip directions of each set of joints and of each planar segment of the excavation surface. or an extreme value selected from the distribution. are intersected with the other half-spaces of the block (i. Consequently. ifjoint i is not a face of the block). for example. Computer programs such as those available from the authors (see the footnote in the Preface). 0110). In subsequent study.. If the actual locations of the traces of joints on the excavation surface are known. log-normal or negative exponen- tial) in which case the volumes and shapes of key blocks would be stated prob- abilistically.7. it is 1 if the block is formed by an intersection with I.. descrip- tive of each joint set in turn. the input can be specified deterministically. A block with parallel sides offers fewer allowable sliding directions than a block lacking parallel sides and therefore proves more stable against sliding. l. As defined in Chapter 4. The first stage output is then a list of half-space codes (e. such as (3) of Fig. For deterministic analysis. 6. joint blocks do not yield key blocks of excavations.. and the ith digit is 3 if both lJ. (the upper half-space of plane t). but in some cases statistical input would be more natural. and shapes are computed for all convex blocks cor- responding to the potential key-block codes. can then generate probabilistic support procedures for use in design. is 0 if the block is formed by an intersection with U. if the block has parallel faces of set i). The ith digit in D.g.164 Block Theory for Surficial Excavations Chap.

Although general reinforcement of the entire excava- tion. (2) general reinforcement. and subsequent stability analysis for the determined key blocks. large.Basic Concepts 165 requirements for the largest key block of any type. it will also be feasible to treat the danger zones with additional special support. Calculation of the support force and its optimum direction are discussed in Chapter 9. permit economical design of rock reinforcement (Fig. For a particular key-block code. The freedom to change directions need not be extreme.I L support of key blocks. The recognition that there can be a maximum key-block size is very important and will be stressed in this and later chapters. If the designer has freedom to choose the direction and/or inclination of a surface excavation. may well be justified. on a regular pattern. Design The results of key-block analysis. there is a specific set of "danger .{ Figure 6. The difference is connected with the kinematic control of failure by the system of blocks. analysis of the key blocks will permit an optimum choice to be made. sometimes only a 10 to 20" shift in the strike ofan excavated slope will greatly reduce the number and severity of the key blocks. This is probably the most significant application of block theory. The extent to which support is actually required for a surface excavation depends greatly on the direction and inclination of the excavation. 6. Figure 6.8 Rock reinforcement: (l) . sudden changes in the degree of stability can be realized as the key-block types and the danger . even though the input data are selected from a statistically defined data set.8). As the direction or inclination of the slope is changed. zones" in the excavation.9 shows a railway cut in which the same rock mass requires different slope angles to be self-supporting on the two sides of the rail line.

.10 Tapered block. 6. let us assume that the excavation is made by a single plane. Occasionally. rock slopes are greatly different from soil slopes. In this respect. pointed into the rock mass. In the latter the factor of safety varies continuously as the slope inclination is changed.166 Block Theory for Surficial Excavations Chap. GOTTIDITIOIUS FOR REMOVABITITY OF BTOCKS IilTERSECTIITIG SURFAGE EXCAVATIOilS In this chapter we consider only those blocks that intersect an excavation surface. as shown in Fig. To start.9 Asymmetric excavation. with normal 0. the factor of safety moves discontinuously from one function to another as the key blocks shift from one block code to another. Assigningthe subscript Figure 6. By P. 6 oo i+i Figure 6. Pr.(Or) we mean "the half-space of plane P. zones shift. a steep slope in rock proves more stable than a flatter slope. In blocky rock slopes. that contains vector 0r".10.

(-0. Chaptet 4.11 Diagrammatic representa- tion of requirements for an infinite.@t) (6. for all space. When (6. Alternatively." the criterion for a block to be infinite is JPNEP*@ (6. Denoting the empty set by the symbol @.6) .s) Finite btocks. the excavation pyramid. By the theorem of block must have a nonempty block pyramid (BP). the excavation pyramid. that is. EP : P. plus the space pyramid. an infinite Infinite btocks.1a) and SP:L. BP: @ (6. are related by EP: U. SP.otLt (6. and space pyramid. SP.) Most of the blocks intersecting a free surface are infinite or tapered. accounts Figure6. convex block.a) is true. BP* @ (6. To be finite.1b) and SP : P. a convex block with one face in the free surface must have an empty block pyramid. JP lies only partly in SP. by the finiteness theorem of Chapter 4. EP.orU. finiteness.3) where n means "intersected with. meaning that it is not contained in SP: JP+SP (6.Conditions for Removability of Blocks lntersecting Surface Excavations 167 i to the free surface. EP.4) As shown in Figure 6.!1. a planar diagram (not necessarily a stereographic projection).2) Since BP:JPNEP (6.

3). Note that the block would be finite with the joint planes alone (i. Equations (6. According to the previous discussion.7) and (6.12). 6.8) are equivalent. meaning that it is completely contained in SP (Fig.8) Equations (6. The conditions for removability of a block are. 6 Introducing (6.7) and (6.5).5) and (6. The condition for a finite convex block to be non- removable \ilas established in Chapter 4. a finite block satisfies (6. This demonstrates the validity of equation (6. or JPcSP (6. as are (6. a convex block must be finite and removable (Fig. JPnEP:@ (6. therefore.12 Diagrammatic representa- tion of conditions for a finite.7) and (6. 6..8). To be a potential key block.9). Tapered blocks. JP+ @ and JPnEP:@ (6. its block pyramid is empty. A tapered block is diagrammed in Fig.10. Since it is finite. whereas (6. a potential key block must not satisfy (6. convex block. without the excavation surface). Block Theory for Surficial Excavations Ghap.7) Equation (6. Thus tapered blocks satisfy both JPnEP:@ and JP: fr (6.10) or JPcSP . Removable blocks.13).4) and (6. while the condition of nonremovability dicates that its joint pyramid is also empty.e) Equation (6.8) are better suited to stereographic projection solution. Figure 6. 6.9) is also the condition for the finiteness of a joint block.9). since it must not be tapered.4) prove more convenient for vector solution. Moreover.e.7) is true if and only if JP lies entirely outside EP.

For these examples we again use the joint data of Chapter 5. Thus it is possible to use stereographic projection to determine the infinite. Both U. and therefore occupies a region between arcs of gteat circles in the projection. The system of joints and free surfaces is listed in Table the area inside circle i. tapered. IDENTIFICATION OF KEY BTOCKS USING STEREOGBAPH IG PROJECTION Stereographic projection offers a direct. is the area outside circle r.13 Potential key block. and removable block codes corresponding toaparticular system of discontinuity and excavation planes. arc half- spaces. and removable blocks. graphical solution of the equations above. A series of examples will apply the criteria for infinite. It will be recalled that both joint sets and free planes are projected as great circles.1 Joint and Slope Orientations for Example Problems Joint Set or Slope Plane Dip.ldentification of Key Blocks Using Stereographic Projection Figure 6. tapered. its code is 1. and L. (J. B (Free Surface) (dee) (dee) I (joint set) 75 80 2 (joint set) 65 330 3 (joint set) 40 30 4 fioint set) 10 270 5 (free surface) 60 0 6 (free surface) 80 90 . In the lower-focal-point projection.1. its code is 0. a Dip Direction. Z. TABLE 6. JP (if not empty) is an intersection of joint half-spaces.

Therefore. and therefore each region can be assigned a code. infinite blocks satisfy the criterion that JP + for any region that is not entirely contained inside the circle SP. . Block Theory for Surficial Excavations Chap. Each of these is simultaneously inside or outside every one of the joint-plane circles. EP corresponds to the area outside circle 5. In other words.14 Stereographic projetion of data of Table 6. 6 A Slope Formed by a Single Plane In Fig. EP Ls : : and SP Us.1. 6. On the figure. the regions of intersection of joint-plane circles have been labeled with appropriate half-space codes.14. the fourjoint sets are projected as great circles together with the gteat circle for a free surface corresponding to plane 5.14. 6. This is true face (5) q" T Figure 6. In Fig. a lower-focal-point projection. while SP corresponds to the area inside circle 5. According to equation (6. The rock mass is in the lower half-space of plane 5. the four joint-plane circles intersect to create regions each of which corresponds to a JP. with free surface of plane 5 only.5).

6. the infinite blocks correspond to joint pyramids with codes 1111. Thus two half-space intersections are not represented on the projection. 1011. 6. Inspection of the stereographic projection of the four joint planes shows that there are only 14 regions of intersection of the four joint-plane circles. By the process of elimination. I . each of which creates two half-spaces. The tapered blocks satisfy the criterion JP @. As shown in Fig. We will consider first the case where the rock mass is the intersection of the lower side of plane 5 and the upper side of plane 6. Gonvex $lopes The stereographic projection also offers a solution for the infinite and removable blocks of slopes formed by more than a single excavation plane. and 1000. these are found to be those corresponding to codes 1101 and 0010.1. The second requires that a removable block have a JP that falls entirely inside the circle corresponding to plane 5.14 establishes the codes meeting this criterion as 0011. this intersec- Plane 5 Figure 6. Every JP with a code shown on the stereographic projection is nonempty and therefore satisfies the first criterion. By inspection. This means that the joint - pyramids of tapered blocks do not appear in the stereographic projection (because an empty pyramid exists only at the origin. 0111. 0101. 0100. Thus the tapered blocks are those formed with joint pyramids corresponding to 1101 and 0010.15. and the origin is excluded from the stereographic projection). 0110. There are four joint planes. 0000. 1100. so the number of possible joint-plane intersections is 16. The removable blocks satisfy the criteria JP * @ and JP c SP. 1001. viewed along the intersection of Ps and Po.15 Convex slope of planes 5 and 6. An inspection of Fig.ldentification of Key Blocks Using Stereographic Projection 171 for plane 5. Suppose that a surface excavation is formed by planes 5 and 6 of Table 6. 1010. and 0001. 1110. Potential key blocks are therefore those whose joint pyramids correspond to one of these three half-space codes.

6. which lies inside SP. In general. and (J5.. The addition of excavation planes does not change the JP codes since they are defined exclusively by the great circles ofjoint planes. .16 Stereographic projection of data of Table 6. then -(l n 8): (-A) U (-B) and . 6. if @a1is the space not included in l.e.16.16 are shown by the ruling.EP where " .. *Since SP : .1 with both free surfaces 5 and 6. EP: Lt)Uu and* SP:UreLu where A v B indicates "the union of A and B. 6 tion creates a convex rock surface. The infinite blocks correspond to N I I I \ Bounds of I SP=UuULu I \ \ Figure 6. and (-A) is the space not included in -8.(A v B) : (* A) n (-B)." The limits of SP within Fig. The regions of intersection ofjoint planes have been labeled as in the previous example. together with the four joint planes. These planes are projected as great circles in Fig.172 Block Theory for Surficial Excavations Chap." means "the complement of" and is interpreted here as "the space not included in" (i. "the other part of"). Since the rock mass is the intersection of l.

6. a and b. 1001.0101.0111. The limits of SP are ruled.t7.0100.0000.16 they are found to be 0110. We can use the theorem of nonconvex blocks presented in Chapter 4 to find EP. 6. The tapered blocks are those having JP abS€nt from the stereographic projection.18 shows the space pyramid corresponding to this case. . and 1010. Using the same criteria as in the previous examples. 6. Point a is in convex block . Since the projection of the JP regions is independent of the choice of excavation surfaces.17. the tapered blOcks are the same as in the previous example: 1101 and 0010. 1111. 1100. Point b is in convex block 8(b) with free surface (Js. Figure 6. First choose two points. Then EP(a) : Ls and EP(6) : Uu.ldentification of Key Blocks Using Stereographic Projection 173 JP regions that do not lie completely inside SP. (4'30)' equation Ep : Ep(a) u Ep(b) Therefore' Ep:Lsuua SP is convex.B(a) with free surface Lr. and 1000. therefore. By inspection of Fig. The removable blocks are formed with JP corresponding to regions entirely included in SP. as shown in Fig. viewed along the intersection of Ps and Pe . 0011. *See the preceding footnote. 1110. with the rulings inside SP. It is determined by* SP:UrAL. They are.17 Concave slope of planes 5 and 6. By Shi's theorem. Figure 6. 1011. This concave region is determined by the union of the lower half-space of plane 5 and the upper half- space of plane 6. Concave Slopes Now consider a rock block in a rock slope formed by planes 5 and 6 but with the rock mass in the shaded region of Fig. 0001.

0100. however.18 JPs when there are no repeated joint sets and the SP for a concave slope of planes 5 and 6. 1010. Suppose that the repeated set is joint plane 1. as noted previouslS because the restriction of sliding directions mobilizes an increasing normal stress as a consequence of a small initial motion.0111. Removable Blocks with One Repeated Joint Set Removable blocks sliding between parallel joints of the same set are inherently more stable than those lacking a repeated set.. the 174 . If any one of these JP codes identifies a region on the projection. The possible JP codes are then 3000. The removable blocks are reduced in number to one. SP=Uuft La \ 0011 \ \ \s I . namely 1001. .e. and so on. 1110. 1100. 1101 and 0010).0101. 3001. only the single block 1001 is entirely included in SP. The tapered blocks are the same as before (i. I I I / 011 1 \ 1111 \ \ \ \ \ ) Plane /\6 \ Figure 6. That is.0110. 1011. Sometimes.0011. and 1111. 1000. . 3010. The infinite blocks are 0000.19 shows a convex block between parallel joints. such blocks are critically located and need to be analyzed.0001. Figure 6. .

region must lie along a segment of the gxeat circle for joint plane 1. with EP: Lr. 6. SP=Urr^tLu N Plane 5 (free plane) s-2.s--- N -3100 otgo Plane 6 (free plane) Figure 6.20).In this case.19 Convex block between par- allel joint planes. SP is the region inside the circle of plane 5 (Fig.ldentification of Key Blocks Using Stereographic Projection 176 Figure 6.20 JPs with one repeated joint set and the SPs for convex and con- cave slopes. 3) of Bounds 8t Bounds of SP=UuUL. Let plane 5 be the free plane. .

176 Block Theory for Surficial Excavations Chap. 6

The joint pyramids corresponding to repetition ofjoint
1 are nonempty if
they appear as segments of circle 1. A particular segment corresponds to a finite
block if and only if it lies entirely inside SP. Codes that do not appear as a
segment correspond to tapered blocks and those that are entirely or partly
outside of SP correspond to infinite blocks. In Fig. 6.20, all the segments of
joint plane t have been labeled with the appropriate JP code. The infinite,
tapered, and removable joint blocks for EP : Ls were then determined by
inspection of the figure. The infinite blocks are 3111, 3110, 3100, and 30@. The
tapered blocks are 3010, and 3101; and the removable blocks are 3011 and 3001.
The same analysis can be repeated for joint plane 2 serving as the single
repeated set, then for joint plane 3, and finally for joint plane 4. The results are
presented in Table 6.2.

TABLE 6.2 Identification of Block Types for Any One Repeated
Joint Set and Excavation along Plane 5 (EP : Ls)

Repeated Joint
Set Infinite Tapered Removable

31 ll 3010 3011
31 l0 3101 3001

0300 0310 0301
1300 1301 0311

0t3l 0030 0031
0r30 I 131 1031

0103 0013 0003
0113 1103 1003
11 13

Now we consider a block, having a single repeated joint set, sliding into
an excavation produced by two free surfaces. First a convex slope will be exam-
ined, with EP formed by the intersection of L, and, (Ju, as in Fig. 6.15. Then
SP :
Us U 2.. The JP codes are not affected by the choice of SP, so the list of
tapered blocks is unchanged. But the division of the blocks into infinite and
removable types is changed. Table 6.3 lists the infinite and removable blocks as
determined from Fig. 6.20 with SP (Js U Le. :

ldentification of Key Blocks Using Stereographic Projection

TABLE 6,3 Identification of Block Types for Any One
Repeated Joint Set and a Convex Rock Slope Formed
by Planes 5 and 6 (EP :
Ls o Uc)

Joint Set Infinite Removable

3l l1 301 I
3l r0 3001
3100 3000

1310 0311
131 1 0301

1130 0031
0130 1030
0131 1031

1113 0003
0113 1003
0103 1013

In the case of a concave slope, with EP - Ls U Uu, as in Fig. 6.17, SP is
reduced to the area of intersection (Js I Lu, as noted previously. The list of
finite, remoyable blocks is then shortened to four: 3011, 3001, 1031, and 1003'
When joint set 2 is repeated, there are no removable blocks in the concave slope.

Removable Blocks with Two Repeated Joint Sets

If a block with a single pair of parallel faces tends to be more resistant
to sliding than a block lacking parallel faces, a block with two sets of parallel
faces must be even more resistant to sliding. Although unlikely to slide, we will
discuss such blocks because they are potential key blocks under special con-
ditions. We have seen that a nonempty joint pyramid of a block with one set of
repeated surfaces is an arc of a great circle, as shown in Fig. 6.20. In the case of
two sets of repeated joint surfaces, a nonempty joint pyramid will be represented
by the point of intersection of the arcs of great circles corresponding to the two
repeated joint sets. The code for such a JP contains the digit 3 in two positions.
Figure 6.21 shows all the JP's corresponding to blocks of this type (for the joint
system of Table 6.1).
Let plane 5 be the only free plane, with EP equal to Lr, and SP - Usi
examination of Fig. 6.21then establishes which of the blocks corresponding to
these JPs are infinite, tapered, and removable. The results are given in Table 6.4.




17E Block Theory for Surficial Excavations Chap. G

331 |

/ \
I \

\ f*-- Plane 5



0303 /

i-gts 1330


Plane 6 \

Figure 6.21 JPs with two repeated joint sets.

TABLE 6.4 Identification of Block Types for Blocks with Any Two
Repeated Joint Sets for an Ercavation Corresponding to
Joint Sets Infinite Tapered Removable

1,2 3300 3301,3310 331 I
1,3 3130 3030,3131 3031
1,4 31 13 3013,3103 3003
2,3 1330 0330, 1331 0331
2,4 l3l3 0313, 1303 0303
3,4 0133 0033,1133 1033

Evaluation of Finiteness and Removability of Blocks Using Vector Methods 179

Now consider a conyex slope with EP : Ls A Uu and SP - Us \J Lu.
Since the JP representations'are unaffected by changing the nature of SP, the
list of tapered blocks is the same as in Table 6.5. The list of infinite blocks is
shortened by two and the list of removable blocks lengthened by two because
blocks 1330 and 1313 become removable. In the case of a concave slope with
EP : L, \) Uu and SP : U, ) Lu, the list of removable blocks is reduced to
a total ofthree: 3031, 3003, and 1033.

TABLE 6.5 Direction Cosines of Normals to Joint Planes
and Free Surfaces Listed in Table 6.1; fi : (Xr Yu Zi

I o.9512 0.1677 0.2588
2 0.7848 o.4226
3 0.3213 0.s566 0.7660
4 -o.1736 0 0.9848
5 0 0.8660 0.5000
6 0.9848 0 o.r736


In the preceding section, methods were developed to test for key blocks using
the stereographic projection. For a computation of finiteness and removability,
formal analysis using vector equations can be used. Mathematical procedures
establishing the emptiness of a joint pyramid were presented in Chapter 2. ln
this section we establish an efficient mathematical shortcut that simplifies coding
and speeds the solution.

Table of Normal Vector Signs

Again let us consider the joint sets and free planes with orientations listed
in Table 6.1. The basis directions are as in the examples of Chapter 2: x is east;
y is north; and z is up. The dip angle (a) is measured from horizontal and the dip
direction (/) is measured clockwise from north. Using equations (2.7) we first
compute the direction cosines for the normals to all the planes. The results are
reported in Table 6.5. It will prove useful to introduce a direction-ordering
index /f; defined by
I'l: x n). no]
sign[(n, (6.11)
where sign (F) : (1, 0,
-1) when -F is (> 0, : 0, < 0).
The values of ft computed from Table 6.5 are shown in Table 6.6. Note
that the values of Il depend on the order in which the joints are written down,
which is arbitrary. However, all possible unique combinations of f and j are

nl L GI 7zo
-'. ' * lL

, C6 ,lrJ\*= Z,I'Z

\/n -
(n-")l L I
Block Theory for Surficial Excavations Chap. 6
(7 = n(n-l) l"'ti
-n ---T--
examined in Table 6.6 for all choices of ft. For n planes, there are nC| such
combinations (90 values of It in the present case).

TABLE 6.6 Values oI I'/ : Sigr [(ni X n;) . n*] for the Planes
of Table 6.1


1. 20 0 l I -1
30 -1 01-l
40 -l -10-l
50 I 110
I 60 -1 -1 -1 -l 0
2 3l o 0 -1 1 I
2 4l 0 l0l I
2 5 -l 0 -l -t 0 -l
2 61 0 -l -1 | 0
3 4l -1 00-1 I
3 5 -l 010 -l
3 61 0-1 r 0
4 5 -l -1 0 0 -l
4 61 l0l 0
5 61 -t -l-10 0

We will be able to use Table 6.6 to determine which half-space combina-
tions determine finite blocks; we will also be able to describe the edges of block

Finiteness of a Block

Consider a particular block defined by the intersection of half-spaces as
given by the block code (D"), for example, (1 0 0 1 1 2). Using Table
6.6, the finiteness of any such block will be judged by the following steps.

1. Choose a block code (Dr):
D": (a, a2 q3 a,) (6.12)
2. For the selected D", determine a signed block code (D") obtained by
transferring each element a, of (Dr) to a value I(a), defined by
if a,: g
if a,: 1
I(a,):{:j if a,:2

so that
D, : (I(ar), I(ar), I(or), . . ., I(a,)) (6.14)

Evaluation of Finiteness and Removability of Blocks Using Vector Methods 181

3. Using the terms It/ defined by equation (6.11), form a testing matrix (Tii)
corresponding to block Dt for each unique combination of i and i : 1,2,
3, . . , , n, ard if*i. (7/i) is a row of n numbers defined by the term-by-
term multiplication of Il and l(ar):
(Tti) : (It/ - I(ar), l.]. I(ar), . . ., ry. I@")) (6.15)

Forexample,forD":(I 0 0 I 1 2),withi:Iand7:2,Table6.6
gives II' : (0 0 1 1 D", corresponding to D" of (1 0 0
-1 1); then
I 1 2), is given by application of (6.13) as (-1 1 1 -1 -1 0). Now
multiply the corresponding elements of lt*z and D" to determine
(Trr):(0 0 1 -1 I 0)

Similarly, computing the row matrixTti for every unique combination of i andi
determines the complete testing matrix (T) as given in Table 6.7. g) is the
(10 x 6) matrix of elements for each row i, j and each column k of Table 6.7.

Rule for Testing Finiteness: If every row of (T) includes both positive and negative
terms, the block Ds corresponding to (T) is finite.

According to this rule, block (1 0 0 1 I 2) isfinite since every row of (?),
given in Table 6.?, contains a mixture of *1 and -1 terms. A contrary example
is offered by block 100122. Following the same steps as in the previous example
will yield the same row elements (Z'r) as in Table 6.7 except that eYery term in
the fifth column becomes zero. This results in a matrix (T) whose nonzero terms
are those corresponding to ft from I to 4 in Table 6.7. The second, third, and
eighth rows of this testing matrix lack positive terms and therefore block (L 0
0 I 2 2) must be infinite. [We knew this to be true by the flniteness theorem
of Chapter 4 since (1 0 0 | 2 2) is the joint pyramid for block (1 0 0
1 1 2) and the latter was previously found to be finite.l

TABLE 6.7 Testing Matrix (?) for Block Da: (l 0 0 1 | 2)


2001-l 0
30-10-1 0
4 0 -1 -l 0 0
5 0 I I -1 0 0
2 3 -1 0 0 I -l 0
2 4-1 010 -t 0
2 5 I 0 -1 I 0 0
3 4-1-r00 I 0
3 5110-l 0 0
4 5 I I -1 0 0 0

Block Theory for Surficial Excavations Chap. 6

Rule for Finding the Edges of a Block Pyramid: If a block is infinite, its block
pyramid is not empty, meaning that it has at least one edge. The edges of a
block pyramid correspond to the lines of intersection of planes i and j for which
(Ttt) lacks positive or lacks negative terms. If (TtJ) contains only negativeterms,
one edge of the block pyramid corresponding to (T) is
-ni X rJ : -111. Con-
versely, if (ftt1 contains only positive terms, one edge of the block pyramid
corresponding to (T) is n; X n7 : *fu

Using the rule above, the edges of the block pyramid of Du: (100122)
are vectors parallel to
-f":-rl1 Xtr3

-f,o - -nr X tr+
and -frn - -n3 X tr+
(In the expressions above, Iry indicates the line of intersection of planes i and j.
The normal vectors n, are positive upward; in the case of a vertical plane, the
normal is directed parallel to the direction input as the "dip direction.")

Finiteness of a Block with Repeated Joint Sets

The method outlined above also applies to analysis of the finiteness of a
block with one or more repeated joint sets. In the block code, D", a repeated
joint set is indicated by the symbol 3. Equation (6.13) gives (+t) for the .I code
corresponding to an element a, : 3. Consider block I2O3I2, with the joint planes
of Table 6.1. Using (6.13) the signed block code, D", corresponding to D": (1
2 0 3 I 2) is D": (-1,0, 1, *1,-I,0). Table 6.6, determining values
of It/, still applies to this block since the same joint planes are being investigated.
The testing matrix, (7), is produced as in the previous example: For each row
of Table 6.6, multiply column k by the /cth term of D". (Since planes 2 and 6
are not involved in the formation of the block being investigated, the pertinent
values for i and j are only 1, 3, 4, and 5.) The testing matrix (T) determined by
this multiplication is shown in Table 6.8. Since every row (fii) of matrix (7)
has both positive and negative terms, block Da: (l 2 0 3 L 2) is finite.

TABLE 6.8 Testing Matrix (") for Block Ds : (l 2 0 3 | 2)

3 4 -l 0 0 0 +l 0
3 5+100*100
4 5+r0-1000

j: I. which is entirely negative. corresponding to a surface excayation.. generally.. 1. is 0 or 1 and i:1.n and a.13).9. *1.3. D. (2) tapered.6. TABLE 6. Using (6. parallel to -Ir4: -llr X tr+. The complete testing matrix (fl is given in Table 6. we must multiply column fr by the ftth term of D". Each of these blocks hasn*lfaces. Block 300321 is therefore infinite.. an*1 describes EP. Now we examine the numbers of different types of blocks in the presence of one free plane. Blocks with No Repeated Joint Sets Given that there are n sets of joints and one free plane separating the rock mass from free space.2. Because plane 5 is not involved in the formation of the block. 0. For each row (Z'i) of Table 6. .4. : (+1. 1.9 Testing Matrix (7) for Block Dn: (3 0 0 3 2 1) 2001+10-l 30-10+10-1 4 0 -l -l 0 0 -1 60-1-1+100 3 +r 0 0 +1 0 -l 4+10100-1 6+10-1+100 4 +1 -1 0 0 0 -1 6*ll0+100 6+1 11000 All the rows of T now have both positive and negative elements: except row 1. The codes for such blocks will be expressed. and 6.. 4. Blocks fall into the following categories: (1) infinite.*1 describes the rock half-space determined by the free plane. the applicable rows correspondto i. AS D" : (qt az a3 an an+r) where a.2. THE NUMBERS OF BTOCKS OF DIFFERENT TYPES IN A SURFACE EXCAVATION In Chapter 5 we discussed the numbers of blocks of different types occurring within the rock mass.The Numbers of Blocks of Different Types in a Surface Excavation 183 Now consider block 300321. and (3) removable. we will consider blocks formed from one half-space of each of the n + 1 planes. -1). Its block pyramid has one edge. that is.

- Since the number of JPs is unaffected by the presence or absence of free planes. The question to be determined is how many of these JPs remain when a free surface cuts through.22 z joints cut a free plane into 2n angles. none of which is repeated. this result is valid here as well. . Block Theory for Surficial Excavations Chap. This is the number of possible joint pyramids. It was established in Chapter 5 that the numberofnonemptyjoint pyramids determined by n joint sets. The number of infinite blocks is the number of nonempty block pyramids. K+1 K+1 (a) Angle in free plane (?) Figure 6. The number of all half-space combinations of the n joint planes is 2" since each a. is n2 n + 2. has two possible values. 2. 6 1.

The nonempty JPs that do not intersect the free plane can be counted by subtracting the result in (3) from the result in (2). 6. as shown in Fig. 6. This is the same as for joint blocks as deduced in Chapter 5.23).(n" . 6. Every angle in the free plane is the intersection of the free plane and some JP. As shown in Fig.The Numbers of Blocks of Different Types in a Surface Excavation 185 3.3n + 2)12 (Fis. 4.22(a). 7.3n * 2 joint pyramids that are nonempty and do not intersect the free surface. 6. Subtracting the result in (2) from that of (1) yields the number of tapered blocks as equal to 2o . This yields nz . 5.23). 6.n * 2)12 (Fis. among the JPs created by the n joint planes there are 2n nonempty JPs that intersect the free plane.10. Removable blocks Free space F ree plane Rock mass I ntersecting free plane lnfinite blocks Figure 6.23 Diagram representing conditions for infinite and removable blocks. finite blocks are those whose JPs are entirely included in SP. half are entirely included in SP and half are entirely include in EP (Fig. From (4). The removable. 6. or in other words. Suppose that there is a free plane cutting through n sets of joints. the number of infinite blocks is equal to (nz 4. The infinite blocks are those whose joint pyramids are not entirely included in SP. the joint planes cut the free surface into 2n angles. The results (1) to (7) are summaized in the first row of Table 6.23). Subtracting the result in (5) from the total number of nonempty JPs given in (2). The number of tapered blocks is the total number of half-space combina- tions of the four joint planes less the number of nonempty joint pyramids. Half of these are on one side of the free plane and half are on the other. Therefore.22(b). the number of JPs entirely contained in SP is (nz .n * 2). .

l)l n(n .r-?\ n2-3n*2 -z. 1X2z-3 . l'l Any 2 repeated sets n(2n .2) n)2 2 selected repeated sets 2n-2 | Z.Z I n>3 n(n --l)2-z n(n . 0 nlm)3 (m>2) .l) n.2-^ 0 Cf.-2 .(nz-z-. 2@ .10 Numb€r of All Combinatiols Numbec of Number of Number of Numb€r of R€peat€d of Haf-Spac€s llf[ite Tapercd Renovable Joint Sets (All JPs) Blocks Blocks Blocks Condition Orepeatedset 2n nzllt*Z )n. z"-(42-n+2) n>l I set 2o-r n -f selected repeated zo-r .1) n23 I 2 2 z selected repeated sets 2o-^ 0 2r-^ n2m) 3 (m>2) Any m repeated sets Cf ..l\ n(n. Zfu .I @ o TABLE 6.z tt>z Any 1 repeatedset n2o-r n2 nlln-t .2"-.

4. each of the joint sets 3 through n. comprising 1 face of the free plane.10. The number of possible joint pyramids is T-' . However. including the free plane. this is also true for joint set . If any joint plane can be repeated' there will be z times as many blocks of each type. and I face each of planes 2 to n. The number of tapered blocks is the total number of JPs.1) angles. 2 faces of the first joint set. Half of these are entirely contained in SP and therefore belong to the finite.2. removable blocks. The number of removable blocks is therefore equal to n . so that blocks all have codes D" : (3 3 a3 a4 an a.Ljoint planes 2 through n will cut plane 1 into 2(n . and the first two joint sets taken twice. The number of infinite blocks is the number of JPs not entirely included in Sp. These blocks have n + 3 faces. Therefore. given in (2). 3. 6. 1. The free plane intersects plane 1 along a line that lies in two opposite JPs. Therefore. The number of all combinations of joint half-spaces is now 2"-2. all nonempty joint pyramids must intersect this joint plane and each nonempty JP is therefore an angle subtended at the origin in plane 1.+r). All JPs must be in the intersection of both half-spaces of joint set 1 and therefore must lie along joint set l. Results (l) to (6) for the blocks with one assigned repeated joint sets are summarized in the second row of Table 6. Since the first joint set is repeated. each of which can be considered to be a nonempty JP.1) taPered blocks. Blocks with Two Repeated Joint Sets Consider that the first two joint sets are repeated. This is the total number of nonempty JPs as given in (2) less the number of JPs included in SP as given in (a). as summaizedin the third row of the same table. the total number of JPs that intersect the free surface is 2. The n . The possible blocks all have the code Dr: (3 o2 a3 an dn*r). 5. there ate 2"-r .The Numbers of Blocks of Different Types in a Surface Excavation 187 Blocks with One RePeated Joint Set Now assume that the first joint set is repeated. 2(n . Each of these blocks has n * 2 faces. given in (1) less the number of nonempty JPs. This yields the number of infinite blocks as equal to n. The number of JPs that are nonempty and that do not intersect the free plane is the number of nonempty JPs given in (2) less the number that intersect the free surface given in (3). 2.

asmeasured by block volume. Most Gritieal Key-Block Types In the previous sections it was demonstrated that a certain number of block pyramid codes dictate finite. the volume of the largest probable block is dependent on the aerial extent of the free planes . This section considers only the application of block theory to the geometric design of the rock slope. But not all are equally critical. The sixth and seventh (bottom) rows generalize the results for three or more repeated sets.1. The number of removable blocks is the number of nonemptyjointpyramids that are entirely contained in the space pyramid. 6 2. As noted previously. Given a particular key-block code. all other practical con- siderations laid aside. Since only one of these is within the rock. PROCEDURES FOB DESIGNING ROCK SLOPES The design of a rock slope entails choosing geometric properties for the excava- tion and temporary or permanent support measures in the face of real constraints of time. The first step in a practical design problem with a new rock slope is to identify the critical key-block types. This will allow you to examine the con- sequences of a change in dip directions and dip angles of the slope planes. there are C| cases like the one considered and therefore all the number characterized for two assigned repeated sets must be multiplied by (n)(n . involving computation of force equilibrium. removable blocks. 188 Block Theory for Surficial Excavations Chap. Two measures of relative importance of key blocks are their size. When any two joints can serve as the repeated set. so the nonempty joint pyramids must lie along the line of intersection of sets I and 2. These are all potential key blocks of the rock excavation. 2. The question of slope support. can be addressed more efficiently in chapter 9. 5. The number of tapered blocks is the total number of half-space combina- tions of the joints less the number of nonempty JPs and is therefore equal to 2"-2 . 4.I)12. and their net shear force. as measured by the difference between sliding and resisting forces per unit volume. These results are summafized in the fourth and fifth rows of Table 6. It follows from (2) and (3) that there is only one JP contained in SP and therefore there is only one removable block. The number of nonempty joint pyramids is therefore equal to 2. precedent. and land use. abrupt changes in the degree of rock slope stability accompany shifts in either the rock slope's dip or dip direction. the number of infinite blocks is 1. 3.

0. For these two reasons. block pyramids 00111. Therefore. -1).Procedures for Designing Rock Slopes 189 and the joint planes forming the faces of the block. The net shear force per unit volume depends on the friction angles of the sliding faces and on the orientation of the sliding direction. see the analysis of the sliding direction in Chapters 2 and 3. a block of type 0011 is comprised of the free plane and one face each of joint planes 1. Assuming that the rock mass isbelowplane S. therefore making an acute angle with the direction of the resultant force (0.e. The manner in which these factors can be weighed in ranking key blocks is illustrated by means of an example. Then the key blocks are also these three. and 0001 (i. Sp is the area inside the great circle for plane 5. t0Ott. The other key blocks have no vector as steep. 1001. blocks of class 00111 are the largest expectable key blocks of the three. a Dip Direction Plane (dee) (dee) Relative Extent I 75 80 Large 2 65 330 Large 3 40 30 Large 4 10 270 Small 5 (free 60 0 Large plane) Figure 6. (Also.11 Dips and Dip Directions of Planes consiilered in Example Dip. because each of these JPs contains vectors in the lower hemisphere..1.2. The only removable blocks are therefore those corresponding to joint pyramids 0011. The other two key blocks involve plane 4 also and lhis plane has limited extent. the steepest vector in 0011 plunges steeply and therefore has a relatively large net sliding force per unit volume. Each of these planes has a relatively large extent.24 is a lower-focal-point stereographic projection of the four joint sets and the free plane (plane 5). Moreover. Assume that the resultant forces are due io gravity only. In other words. so the block formed by them can run to large size. but it can be assigned a relative "extent". and vice versa.) The question is: Which of the three key blocks is most critical? JP 0011 is a spherical triangle meaning that it has only three joint faces. A joint plane's probable maximum area cannot be measured. and 00011). the dips and dip directions of the planes are listed in Table 6. joint sets that tend to have long traces on exposed surfaces are assumed to have a large aerial extent.11. blocks of code 00111 are more critical than either of the other' . whereas JPs 1001 and 0001 are spherical rectangles having four joint faces. For convenience. This will be discussed further in Chapter 9. The last column of this table states an estimate of each plane's relative extent. TABLE 6. Consider again the four joint planes and the first free surface (plane 5) of Table 6.-and3.

190 Block Theory for Surficial Excavations Chap. 6

\ Free plane

ol 11

Figure 6.24 Lower-focal-point stereographic projection of data in Table 6.11.

stereographic Projection construction for the Limiting slope
Angles When the Slope Direction ls Given

We have now established which blocks are potentially critical for a particu-
lar rock slope. Suppose that the slope is fixed in direction but there is freedom to
adjust the slope angle. This is frequently the case in engineering for transporta-
tion routes. For every potential key block, represented by a JP, and the assigned
strike for the rock cut, it is possible to construct two great circles serving as
envelopes to extreme vectors of the joint pyramid. These envelope circles bracket
the range of slope angles for which blocks corresponding to the enveloped JP
are removable. To perform this exercise it is necessary to construct a great circle
passing through both: a given point anywhere on the projection plane;and a
given point on the reference circle (the strike vector for the cut).
To construct a great circle with given dip direction (given strike) through
any assigned point (see Fig. 6.25):

Procedures for Designing Rock Slopes

..' Reference
!4- circle

Figure 6.25 Great circle with assigned '
sin {6}
strike that passes througlr a given point'

1. Plot points ,4 and -B representing opposite strike vectors for the cut.

and B are the intersection points of the reference circle and a diameter of
the reference circle oriented perpendicular to the dip direction of the cut
2. Loiats the point P through which the great circle is required. P is an
extreme point of a JP.
3. Let R be the radius of the reference circle and let < APB: d (Fig. 6'25)'
if C is the center of the required
Then great circle: 4 ACB : 2 6; and
< ACO: { BCO : d. Calculate the radius r of the required great circle
-------T (6.16)
- sln o

If a is the dip angle of the required great circle,
R (3.e)
Equating (6.16) and (3.9) the dip angle of the required great circle is
a:g0-6 (6.17)

4. Construct the required great circle at center C and with radius r deter-
mined bY AC : r-

Returning to the example of Fig. 6.24, it has been established that JPs of
interest for a cut with dip direction corresponding to plane 5 (dip direction equal
0) are 0011, 1001, and obot. In Fig. 6.26, the method described above was used
to construct greai circles enveloping extreme points of these three
joint pyramids.

192 Block Theory for Surficial Excavations Chap. 6



Figure 6.26 Great circles of assigned strike that just contain three Jps.

These gireat circles dip 57.7",46.3", and 36.4'. If a cut slope is steeper than
57 .7", all three JPs are potential key blocks. But if the cut is sloped at an angle
less than 57.7o,0011 no longer determines a removable block. f
ihe cut is flatter
than 46.3",0001 ceases to define a removable block. And if the cut is flattened
to less than 36.4", 1001 ceases to determine a removable block as well. Table
6.12 summarizes these conclusions. This example shows that rock slope safety,
measured in terms of a ratio of resisting to driving forces, must be a discontinu-
ous function of slope angle. As stated earlier, this behavior makes rock slope

TABLE 6.12 Key Blocks of Each Dip-Anete Interval"

Dip, a Dip Direction, fl Removable
Number (dee) (dee) Key Blocks Blocks

I 0-36.4 0 0000,1000
2 36.446.3 0 1001 0000
3 46.3-57.7 0 1001,0001
4 57.7-90 0 1001,0001,0011

tSee Figure 6.26.

Procedures for Designing Rock Slopes 193

stability analysis distinct from that of soil slopes. With respect to the example,
Table 6.L2 implies that there is no change in the safety of the rock slope with the
assigned strike when the dip angle of the free surface is steepened from 60o to
90". Such a conclusion is particularly valuable when choosing a construction
method for the rock excavation, for it infers that steep benches can be used.

Use of Vector Methods to Calculate the Limiting Slope
Angles When the Slope Direction ls Given

The problem solved in the preceding section was to find the dip of a cut
slope, given its dip direction, such that a given JP ceases to determine a remoY-
able block. A solution can also be obtained using vector methods. The procedure
will be to compute the edges of the JP and for each edge to compute the dip of a
free plane containing both the line of strike of the free plane and the line of the
edge of the JP. The edges are determined as lines of intersection of the joint
planes, as discussed previously [equation (2.11) and Example 2.3)J. The solution
will be worked out for the joint sets of Table 6.11 and free planes dipping north
(i.e., striking east-west).

1. Compute the unit normal vectors fi, : (X,, Y,, Z,) for each joint set, with
equation (2.7).The computed values of fi, arc given in Table 6.13.

TABLE 6.13 Unit Normal Vectors

i Xi

I o.95t2 0.1677 0.2588
2 -0.4531 0.7848 0.4226
J o.3213 0.5566 0.7660
4 0.0000 0.9848

Compute the line of intersection f,, of each pair of planes, using equation
(3.7), and convert each to a unit vector, I,r.The results are listed in Table

TABLE 6.14 Lines of Intersection lii : (Xij, Yii, Zii)


Xij Yti Zij

2 0.8378
-o.1347 -0.5289
J -0.8049 0.5930
4 0.1658 -0.9857 0.0292
3 o.464r 0.6125 -0.6398
4 0.8895 0.4291 0. I 568
4 o.t66l -o.6282 -0.1350

194 Block Theory for Surficial Excavations Chap. 6

3. Ifthe cut planes all have the same dip direction, the strike line, ,S, is a
common line of intersection of all the cut planes. Since the x axis is east,
: -S
,S (1, 0, 0). The plane that contains b"th ;; 7,;;;; normal n,,
calculated bv
n;J:Sx1,, (6.18)
Let Pq be the plane having fi,, as its unit vector. The equation of p,,
Aux I B,tY + c,fi: 0 with the parameters as listed in Table 6.15.

TABLE 6.15 Parameters of planes ps7 Containins
4,, and (1, 0, 0)

Dip,a Dip Direction, p
Au Bu CiJ (dee) (dee)

I 2 0
-0.8456 -0.5338 57.73 0
I 3 0 - 0.5931 36.38
-0.8050 0
I 4 0
-0.0296 1.69 0
) -0.9995
3 0 o.7223 0.6915 46.25 0
2 4 0 .-0.3433 o.9392 20.07 180
3 4 0
-0.2102 -0.9776 12.13 0

4. Having computed a series of planes containing each line of intersection,
the next step is to choose each one of these as a fiee plane, in turn and
the methods of this chapter or of Chapter 2, findittr t.y blocks of each
P,v' only three of the Prr's generate key brocks i prs, prr, andp,r.
The solu-
tion for the key blocks of each of these free planes willgenerate Table 6.12.

stereographic proiection construction for the Limiting slope
Directions When the Slope Angle ls Fixed

In the preceding section it was assumed that the slope direction was given
and it was possible to adjust the slope angle. Here we examine the alternative
case where the slope angle is prescribed and the cut can be shifted
in direction.
The problem posed corresponds to constructing a great circle of determined
radius through a given point. In Fig. 6.27,1 is the poioitnrough which the great
circle is desired. The radius of the reference circle is n. the dip angle of the
slope is given as c, the radius of the great circle for the cut slope cuo t. calculatej
from equation (3.9):

r: R-
cos d

To draw the great circle, it is convenient first to locate point
4 the opposite of
point A. As shown in Fig. 6.27, point B is located at distance frfrom the center
of the reference circle along the extension of line AV and, from (3.7):

Procedures for Designing Rock Slopes

Figure 6.27 Great circle of assigned inclination that passes through a
given point.

The required great circle can be drawn with radius r from point C, where
CA: 6: r.
With the construction above it is possible to draw a great circle through
the extreme points of joint pyramids corresponding to potential key blocks. As
the direction of the cut is shifted through these limiting orientations, the key-
block type changes from removable to nonremovable, or vice versa. Consider
JP 0011 in Fig. 6.28. The corners of this JP are lrr, I"r, and lrr, which are the
projections of i, z, tzt,and f,3, respectively. Given that the angle of the cut is 60",
with dip direction 0o as listed in Table 6.11, we can construct great circle P,
representing the plane of the cut slope. As noted previously,JP 0011 then belongs
to a potentially critical key block. Now, using the procedure outlined above,
great circle P,, is constructed to pass through.I12. With a shift in orientation of
the cut slope to that of plane P t z, JP 001L ceases to determine a removable block.
A further shift to Pr, (Fig. 6.29), passing through lrr, drops JP 0001 fromthe
list of removable blocks. The dip direction of the original cut slope design was
360". Turning it a mere 13.3" to direction346.7 (that of Prr) Fig. 6.28 deletes key
block0011. Turning it24.1o to dip direction 335.9 deletes keyblock 1001 (Fig.
6.29). We see that slight adjustments in orientation may significantly enhance

r96 Block Theory for Surficial Excavations Chap. 6

// tu)t,
l \\

\jt\ I


0t 11


Figure 6.28 Great circle of assigned inclination that just contains Jp 0011.

Use of Vector Methods to Find the Limiting Slope
Directions When the Slope Angte ls Given

The stereographic projection solution presented above has a parallel
computation using vector methods. It is required to find the dip direction of
plane passing through a unit vector a : (A, B, c) when the dip an gle,
6,, of the
plane is given. Let the unit normal to the desired plane be ff : (i, y,
z). The
system of equations to be solved is

AX+BY+CZ:O (6.1e)
(x, Y,Z),(0,0, 1) : cos d (6.20)

X'+Y2+22:l (6.2r)
and A'+B'+C2:l (6.22)
In the above, A, B, C, and a are known and it is desired to determine the
components of i6.

1can be calcu- lated.14). given unit vectol I. For each intersection line formed by the system of planes. 6. l I Ipn I I I \ / brot . 6. Then the dip direction has been calculated from the components of the normal.Cz)t/21 6'23) and Z:cos& Using equations (6. the equation of plane P. and for a. C) and dip a. The solution.16. can be computed.23). / / / I \. c2)t/21 I Y:i7zl-BccosdT.. The first plane on Table 6. The intersection vectors (t) are computed as in the previous example (Table 6. Then the dip direction of P. The results for dip direction agree with those determined previously using the stereographic projection.28.// ./ /. The results are stated in Table 6. given in the appendix to this chapter.16 is Pr2 of Fig.t : (A.Procedures for Designing Rock Slopes 197 .29 Great circle of assigned inclination that just contains JP 0001. is I l{ : T+tl.23).':'/ Figure 6.: ffio. the components of the normal to the required great circle have been calculated using (6. B.29.-AC cos d :E .4(sin2 d.B(sin2 e . . The second is Prt of Fig.

Using the list of critical JPs.98 4 0.4462 o.5 60 66.3031 0. and then the fourth plane and examining Fij. to Po and the traces are coded by Iine type. is the region within the great circle for plane 5.5 -0.7924 o.8427 -0.5 60 34.5 60 79.83 REMOVABTE BTOCKS IN AN EXGAVATED FACE All of the preceding analysis has been focused on determining the combinations of joint half-spaces that create potential key blocks in an excavation of stated orientation.5 60 328. so the space pyramid. y1.1580 0. we will again use the sets of joints pr. pn and the slope face p. Dip Direction. p Xi Yt Zt (dee) (dee) 2 o. the remainingjoint planes determine a removable block 2001.5 60 -0.69 2 0. Suppose that a map is available showing the lines of intersection (trace9 of actualjoint planes with the excavation face.5 60 33s.50 4 -0.6448 0.5 60 346. These Jps define the removable blocks given the four sets of joints.8566 0. It is impor- tant to search for all the blocks because the one missed might be the one that .1268 98.8514 0. we can analyzethetrace map to locate all the removable blocks and all combinations of removable blocks. third.r992 0. SP.88 4 0. 1001. The rock mass occupies the lower half-space of plane fr.5 60 4r. 21. oi Table 6.8563 0. The Jps that are entirely included within the SP are 0011. 198 Block Theory for Surficial Excavations Chap. Simitarty deleting in turn the second. Whether or not a key block actually exists once an excavation is made depends not only on the orientations of the individual joint planes but on their positions in the face.42 3 o.55 3 o.14 we can also find removable blocks corresponding to any subset of three joints. To demonstrate the method of delineating the actual removable blocks. When plane p1 is deleted.16 components x1.21 4 *0. 6. 6. and 0012 as detei- mining removable blocks.4r 4 0. and.7746 60 296.5 -0.3492 0.7422 0.6001 -0. 0021.5 -o.3871 0.8112 60 159. All of the joints in Fig.L4 fot JPs entirely included in the SP yields JPs 1201. From Fig.30 belong to one or another of the sets P. Figure 6.87 3 0.7108 0. or Normal unit vector fu1 ofthe prane Containing fll anilHaving Dip Angle c Plane a Dip. pr.4947 0. 6 TABLE 6.1292 60 261. and 0001.5 -0. How can the key-block analysis be applied to identify polygons belonging to key blocks ? We will now demonstrate how to solve this problem.11. 6.5781 o.3s38 0.7904 -0.48 3 0.6243 60 223.86 4 0.30 is a geological map of slope face Pr. pr.

The following procedure will locate all the removable blocks of the face: L. Assign Code: I 0 I 2 2 I 0 2 3 0 I 2 4 0 I 2 0001 generates map code (MC) 0101 for a map of plane Pr. This direction. This suggests creating a special map code (MC) corre- sponding to each of the critical JPs as shown in Table 6. In Fig. . direction 6 is r. Assign Code: Code2. Establish the downward direction of the geological map (FiS. respectively. and 4. 6.Removable Blocks in an Excavated Face Map code JP ru:0121 :r ZtZ: 0101 -t\\\\\t. initiates a progressive collapse of the entire face. is the dip of planePr.esented by point D on circle 5. 3.14 we can determine that point D is inside circle 2 and outside circles 1. the upper half-planes of joint traces 1.1 I .3. shown ur 6 in the margin of the ffi&p.30).p. 3.14.18 shows the Jp codes and the MC codes for all the removable blocks in this example. For example. Therefore. But the upfer half-plane of joint trace 2 on the map corresponds to the lower half- space of joint plane 2. In Fig. Table 6. 6. Assign Code: l.30 Geological map of traces of joints as seen in plane 5 of Table 6. and 4 as seen on the ffiap.17 Map Codes (MC) for Face Ps Corresponding to Code Corresponding to Code Corresponding to Plane 0. 2101 F--1 : 1201 -= IT]TTTIM : 0112 llllllllll JOrnt 1 2 3 4 t D Figure 6.17. and 4. correspond to the upper half-spaces of joint planes 1. 6. JP TABLE 6.

A removable block zorrc with three joints is a line-segment loop that satisfies the following conditions: . The map codes are those that lack any 2's. segment cD is along joint set 3. Since all segments satisfy condition (b). loop ABCDEFA of Fig. Pr. Locate the removable block zones corresponding to critical JPs with three sets of joints. Corresponding to MC 1. Segment FA is along joint set 2. MC is 0. the zone is below the segment. 1201. which it is. Locate the removable block zones corresponding to four sets of joints. the MC is 0 and the zone is above the segment. Segment BC is along joint set 1. is 0 and on the lower side of the joint trace if the map code is 1. It is. and 2101. The map codes for these JPs are 0112.18 JP MC Number of Zones in Joint Pyramid (Map Code) Fig. A removable block zone is a closed line-segment loop that satisfies both of the following conditions: (a) It consists of the joints of all sets P. of the loop the loop area is on the upper side of the joint trace if the map code of P. 6. and p4. 3. Segment AB is along joint set 4. . so the zone must be on the upper side. . the zone constitutes a removable block. The segments of this loop are as follows: . For example. . which it is. with MC 1. Segment EF is alongjoint set 3. 6. 6 TABLE 6. 0111.012L. . Segment DE is along set 2 and MC is 1. . (b) For any point P. . pr.200 Block Theory for Surficial Excavations Chap. They are 0101.30 is a zone corresponding to map code 0101. the zone must be on the lower side of the segment.30 0001 0101 3 0011 0t l1 0 1001 1l0t 0 0012 ott2 2 0021 ot2l 2 1201 1201 2 2001 2lol 2 2. so the zone must be below the segment which it is. The MC is 0 and the zone must therefore be on the upper size of the segment in order to satisfy the condition for a key block. and 1101.

Segment HI is along set 4 and the MC is therefore 1.(0. Y.30 remain in place. which it does.1e) (x.. of the loop. All segments of the loop satisfy condition (b). 6. For example. so the zone must lie above the segment. 6 Appendix (a) It includes segments from any three of the four joint sets. These numbets can be checked against Fig.22) AX+BY+CZ:0 (6. z). 1) : cos d (6. Examination of Fig. The zofie must be on the lower side of the segment. APPENDIX soturroN oF slMUtrANEous EOUATIONS (6.0. Segment /Jis along joint set 1 and the MC is 0.30 with respect to map code 0121. Segment "/i( is along joint set 2 so the MC is 1. According to the assumptions of block theory. which it is. consider loop GHIJKG of Fig. Table 6.30.19) TO (6. Procedures for kinematic and stability analysis are presented in Chapter 9. which it must to be a key block.20). Segment KG follows joint set 4 and the MC is therefore 1. the zone is above or below the segment corresponding to map codes 0 and 1. (b) For any joint set P.Chap. will maintain each of these blocks. the entire face will be stable if all the zones detimited in Fig. 6. 6. However. Since the zone lies below the segment. lt:'"* | (1) . respectively.20) X2+yr+22:l (6. 6.18 gives the number of closed loops that satisfy the key-block conditions for each segment for all critical JPs. it satisfies the condition to be a key block. One way to assure this is to establish a rock reinforcement system that.30 establishes the following properties for the zone: Segment GH is along set 1 and the MC is therefore 0.2r) and A2+82+C2:L (6. The zone lies below the segment. at least. some or all of these blocks may in fact be stable because of their friction properties or their attitude relative to the free space.22) From (6. so the zone mustlie above the line segment in order to satisfy the condition to be a key block.

. we have (A'+ Bz)Xz +QACcosd)X+ C2 cos2d . Az + 82:1.v_. .C1 (6) Substituting (6) into (2) gives us Y:+ulrycosd+AN/wE=ef (7) From (6.C2. . Ar+Cz_l:_82 Substituting this in (7) gives Y:*reBccosd +AN/W@ (8) Equations (1). (6). 6 Combining (1) with (6. AC cos d + BN/sin-' d. .23) given as the solution.22). and (8) constitute equations (6.22). Bz sin2c:0 (4) The solution to (a) is I X:plVtl-Accosn +B (s) From (6.2OZ Block Theory for Surficial Excavations Chap.Insertingthis in (5) gives r_ X : fttl.E(-C cos q.19) gives us I .AX) (Z) Combining (1) with (6.21) yields Xz + Yz:1-cos2d:sin2d (3) Then substituting (2) in (3) and expanding.

203 . chemical. ship docks. triggered by loss of key blocks. ware- houses. constant temperature and humidity. In a self-supporting underground open- ing. power generation. and even aircraft hangars. nuclear waste. and other purposes. compressed air. underground space may also be more economical than space for equivalent uses aboveground. may create expensive stability problems. or thermal loads. practical use can be made of the principles to be demonstrated here. If the rock cooperates. and high resistance to physical. industrial operations. defense. sports facilities. particularly as regards orientation of long excavations. Chambers are mined to store water. and arrangements of openings to minimi ze the danger of block movements. or earthquakes. and other com- modities. Underground space is being used increasingly for storage. oil. Complex arrangements of openings create space for offices. shapes. Block movements. the initial state of stress will be concentrated by the opening and flow around it in such a way as to preserve the closure and interlock of the joint planes. Advantages of underground space include proximity to the need. Block theory provides optimum choices for the orientations. Since it is often feasible for the designer to make adjustments to the layout. rock crushers. chapter 7 Block Thoory for LJnderground Chambers This chapter shows how to apply block theory to engineering for underground chambers. mining. in turn. landslides. storms. power plants. This depends largely on the joints and the system of rock blocks. promote opening of joints and looseoiog. invulnerability to attack. One way to realize economy in development of underground space is to choose an arrangement for the three-dimensional network of excavations that requires only minimal artificial support. which.

P(a) divides the whole space into two half-spaces denoted U(a) and Z(a). The following chapter. edges. Therefore. The elements of these openings are planes. essentially prismatic rooms. entries. and cylinders.let 2 : (0. This chapter shows how to determine the key blocks of all types of details formed by the intersection or union of planar excavation surfaces. FLOOR. U(-a) : L(a) and U(a) : L(-a).000 cubic meters of water. most contain some of the features shown in this example. 7 Figure 7. devoted to tunnels and porta. . (From Broch and Odegaard. branches or bifurcations. bends. with permission. enlargements. including large. In the following discussion . adds the complexity of curved excavation surfaces.1 Layout of a rock cavern to store 20.1 shows a plan through an underground water storage complex in volcanic rocks. 1) be the unit vector pointed upward. corners. AND WALLS Let t be a vector and P(a) be the plane normal to a. 0. KEY BLOGKS IN THE ROOF.204 Block Theory for Underground Chambers Chap. Although every underground project is unique. and intersections.) Figure 7. Vector a points into U(a) and vector -a points into Z(a). pillars.

that is. Let wall i be determined by its unit normal vector r?. In this and subsequent figures of this chapter. the appropriate stereo- graphic projection is shown in Fig. The criteria governing removability of a block now dictate that its JP plot in the region above the great circle for the plane of the floor.1) where SP : L(2) and @ means "empty.3 shows a rectangular underground room. .The criteria for removability of a block are stated in equations (6. If the resultant force is given by weight only.2(b). EP : L(2) and SP : U(2). the region corresponding to EP will be shaded while the region corresponding to SP will be left unshaded. EP : U(2) and SP : L(2). and Walls Removable blocks in the roof. (b) a horizontal floor. A block is removable in wall i if and only if its JP projects inside the unshaded region of the stereographic projection for wall i. such a block can never be a key block because every vector inside such an SP is upward.2) For each of the four walls of a rectangular opening. a potential key block. Any JP that is entirely included in the unshaded region determines a removable block and. so that EP : L(fi.2(a) presents a lower-focal-point stereographic projection of a horizontal plane representing a roof. Since the rock is in the lower side of the floor plane. Figure 7.10).3. 7. Floor. Removable blocks in the floor. Since the rock is found on the upper side of the roof plane. Since the walls are vertical planes. For a block to be removable in the roof: JP+@ and JPcSP (7. 7. Removable blocks in the walls.Key Blocks in the Roof. Figure 7. therefore. point- ing toward the free space.2 SP and EP for: (a) a horizontal roof. inside the great circle shown in Fig.) and U(fr') SP : (7." Figure 7. SP is the region below the roof. they each project as a straight line along a diameter of the reference circle parallel to the strike of the wall. and therefore outside the great circle for this plane.

Figure 7. BLOCKS THAT ARE REMOVABTE IN TWO PTANES SIMULTANEOUSLY: GONGAVE EDGES A block that is removable simultaneously in two faces of an underground excavation can become very large and still fit inside the excavated space. Block Theory for Underground Chambers Chap. 7 Wall 1 Wall4 Wall3 Figure 7. A block that is simultaneously removable from two adjoining planes will be said to be "removable on the edge" of the two planes. Consider E1r.4 shows the four vertical edges formed by the lines of intersection of the four walls of a prismatic gallery. For a prismatic excavation. Any block that is removable on this . and Wu.3 SP and EP for the walls of a prismatic underground chamber. Wall/wall edges. that is. such a block is potentially very dangerous. or a wall and the roof. We will consider first concave edges. respectively. walls W. those in which the rock mass is concave. it is possible to choose an orientation that minimizes the risk of encountering such a key block since the number of key blocks of this type is limited. belonging to the intersections of two walls. formed by the intersection of walls 1 and 2. In the following. the roof and floor will be called. a wall and the floor. There- fore. there are 12 such edges. However.

4) For block 2.5(a) shows an example of such a block. block 1. Block I is contained in the 2 is contained in the side of wall 2.3) and BPr: L(fr') n JP: g (7.. @ (7. For -r0. edge is concave. EP2 : L(frr) (7. EPr : L(fr t) (7. block / and block 2.4 SP and EP for the wall/wall edges of a prismatic underground chamber.6) .Blocks That Are Removable in Two Planes Simultaneously: Concave Edges Figure7.s) and BP2: L(frr) n JP . Using Shi's theorem (Chapter 4) it is possible to view a concave block of this type as the union of two convex blocks. and block -r?. Figure 7. side of wall 1.

The criteria of removability of a block containing edge 1 are JP+ @ (7.208 Block Theory for Underground Chambers Chap.8) From Shi's theorem for nonconvex blocks (4. SP is convex.uEP. denotes "the other part of" or "the complement of.nSP.11) and SP2 : U(frr) (7. (b) roof/wall. (7.10) tt tt where .(B U C): (-B) n (-C) and -(^B n C): (-A) U (-C) Then SPl : U(frt) (7.30).5 Blocks intersecting edges of a chamber: (a) wall/wall. ^.r2) Since EP is concave.7) and EPnJP:@ or JPcSP (7.13) SP: U(frr)nU(tt.z) (7. (7. 7 htr Figure 7." As noted in Chapter 6.r4) . so SP:SP.e) and SP: -EP (7. EP:EP.

. Figure 7.5(b).4 shows the stereographic projection solution for the removable blocks of the four wall edges. 7. A concave block that is removable in the edge between a wall and the roof is shown in Fig. A JP belongs to a removable block in edge E y if and only if it projects entirely inside SP. if an edge is the line of intersection of walls i andn and the space is convex.Blocks That Are Removable in Two Planes Simultaneously: Concave Edges 209 In general.1 and therefore in the unruled area of the appropriate stereographic projection. Such a block must be simultaneously removable in both the wall and the roof. SPir: U(fi.15) Figure 7. Wall/roof edges.)^U(fi1) (7.6 shows a Figure 7.6 SP and EP for the wall/roof edges of a prismatic underground chamber.

. @ ( Figure 7.. EP.18) BP2 : L(fr') n JP .t7) while for block 2.7 SP and EP for the walffloor edges of a prismatic underground chamber. For block 1.t6) and BPr: U(2)^IP: @ (7. and block 2 in the -r0. Consider edge . side of wall 1. EP2 : L(fr t) (7. 7 plan of a rectangular opening drawn in the plane of the roof.Er . which is the line of intersection of the roof (wall 5) and wall 1 . : 9121 (7.21O Block Theory for Underground Chambers Chap. Each remov- able block of E1r is the union of block l in the 2 side of the roof.

Blocks That Are Removable in Three Pfanes Simultaneousfy: Concave Corners 211 For the concave block.6 shows the regions corresponding to SP for each of the wall/roof edges. SPi. the criteria for removability of such a block are JP+@ and JP c SP. consider corner Cr* of Fig. wall 1.) (7. Each block of corner 1 is the union of three blocks: Block 1 in the 2 side of the roof Block 2 in the -r0.9 shows the four floor/walfwall corners. Wall/floor edges. u EP. : U(2) U(ri)) ^ Figure 7.24) and BPl :U(2)olP:@ Q.oo.2s) . First. if a JP plots in the white area of one of these stereographic projec- tions.8. Again.2r) In general. 7. : 9721 (7. The SP regions corresponding to each of these edges is the unshaded region of each stereographic projection. EP.23) where SP. There are eight such corners.8) apply with SP:L(t).oo. Using -rD. all blocks formed with this JP are removable in the corresponding edge.30) gives EP: EP. (7.. formed by the intersection of the roof.r. Shi's theorem (4. The same arguments apply for the removability of concave blocks in the edge of a wall and the floor of an underground chamber. Any block in the edge of wall i and the floor is the union of two blocks: block 1 in the -2 side of the floor plane. Shi's theorem.r. and wall 2.." A concave rock mass is created in each corner where two walls intersect the roof or where two walls intersect the floor of the prismatic chamber considered previously in this chapter. side of wall I and Block 3 in the -rC'" side of wall 2 For block 1.7 shows a plan through the floor of the rectangular opening and the four wall/floor edges. BTOCKS THAT ARE REMOVABLE IN THREE PTANES SIMULTANEOUSLY : CONGAVE CORNERS Blocks that are removable simultaneously in three intersecting surfaces are said to be "removable in a corner.20) and the criteria of removability (7."oor:L(2) n U(lA.U(fir) (7.7) and (?. and block 2 in the side of wall i. for wall i intersecting the roof.22) Figure 7.8 shows the four roof/wall/wall corners and Fig. 7. (7. Figure 7.

since SP: -EP.2e) The criteria for a block to be removable in cornat Cp5 are the same as (7. @ (7.8).9). EP: EPr u EP. but in place of (7. For block 2.7) and (7. u EPt (7.27) while for block 3. 7 Figure 7.8 SP and EP for the wall/wall/roof corners of a prismatic underground chamber.q (7.26) and BP2 : L(fr') n JP . EP3 : L(frt) (7.28) and BP3: L(ftr) n JP .30) Then.212 Block Theory for Underground Chambers Chap. . EP2 : L(fr t) (7.

34) . Similarly.8.. finally. if a concave corner is formed by the roof. 7.) o (-EPr) o (-EPr) : (.oor : L(2) a U(t. plane f. SP: L(2). SP:-(EPl uEP2uEP3) : (*EP.Blocks That Are Removable in Thtee Planes Simultaneously: Concave Corners 213 Figure 7. for a floor/wall/wall corner a block is removable if and only if it satisfies conditions (7.U(2)) A (-L(fr')) n (-L(ftr)) or.30) and (EP') : L(2) (7.r.8) together with (7. and plane 7.U(fr')^U(tnz) (7.31) In general.32) The regions of SP for each of the roof corners are shown in the stereographic projections in Fig. SP.26) to (7.7) and (7.9 SP and EP for the wall/wall/floor corn€rs of a prismatic under- ground chamber.) n UQD j) (7.33) BP1 :L(2)^JP:@ (7.

214 Block Theory for Underground Chambers Chap.2. 7 If the concave corner is formed by the intersection$ of the floor. EXAMPTE: KEY BLOCK ANALYSIS FOR AN UNDERGROUND CHAMBER The principles stated in the preceding section will now be illustrated by means of an example. If SP is convex. Vector methods may also be used to judge if a particular JP is contained in a particular SP. without touching any boundary plane. 7. Dip Direction. Since concave space pyramids arise at intersections of chambers. they will be discussed later in this chapter. Ws 90 I l8r 47x80 Walls Wz. because of their steep dip. there was freedom to rotate the orienta- tion only around a vertical axis.1.r. and wall7.r. be con- tained in SP in order that JP be contained in SP.*.*.r. and 3. all the edge vectors ej are computed for the JP.10. .1. the critical joints of the chamber are joints I. JP belongs to a removable block if and only if each edge vector er belongs to each half-space of SP. The joint sets and excavation planes of the chamber are described in Table 7. wall i. Then a test of each half-space of the SP must be made to see if it contains the e. This combined gate and surge chamber for a hydroelectric power station is approxi- mately a flat. concave edge. SP.. If SP is not convex.9shows the SP regions corresponding to the four floor corners. prismatic box 12 by 47 by 80 meters with its long dimension oriented vertically.10 a Dip. TABLE 7. 7. which are very large.. : U(2) n U(fij) A U(fri) (7. During construction. The stereographic projections shown in this chapter to test whether a JP is contained in an SP offer a simple and direct method to judge the removabil- ity ofany block in a concave corner. Figure7.1 Joints and Excavation Surfaces of the Chamber Shown in Fig.35) and a finite block is removable in the corner if and only if its JP is contained in SPr. W+ 90 28r 12x80 Roof [Zs 0 0 12x47 Floor We 0 0 12x47 aThe direction given is that of the normal to the wall..r. Consider the underground excavation shown in Fig. or a face ofan underground gallery. From Table 7. p Extent Spacing Plane (dee) (deg) (m) (m) Joint I 7l 163 50 I Joint 2 68 243 50 15 Joint 3 45 280 20 l0 Joint 4 t3 343 l2 l0 Walls Wr. First. it is necessary but not sufficient that all e. The critical excavation surfaces are the roof and walls 1 and 3.

TAB. these JPs determine the removable . L(2)is the region outside the reference circle. the JPs that are entirely included in SP for the roof are 110L and 1011. The four sets of joints are projected in Fig. 7.2 Inward Normal Vectors for the Walls Dip of Dip Direction of the Wall Normal (deg) Normal (deg) I 0 118 2 0 28 3 0 298 4 0 208 Key Blocks of the Roof.Example: Key Block Analysis for an Underground Chamber 215 ora6 Wr --\ | E.LE 7. czso In order to apply the relationships discussed previously.i Ji I crlo E46 Figure 7. Floor. it will be helpful to establish the directions of the inward normals to the walls.2.*w. These are given in Table 7.11 and the reference circle is plotted with a dotted line. and Walls : The space pyramid of the roof is SP L(2) (7 .10 Underground chamber discussed in the example.l). Therefore.

12 Key blocks of wall 3. 7 Figure 7. E 1111 Joint 3 E F Figure 7. . Block Theory for Underground Chambers Chap.ll Projection of joint data given in Table 7.r and key brocks of the roof.

and the JPs entirely included within it are 1110.2).Example: Key Block Analysis for an Underground Chamber 217 blocks of the roof. Because the roof and the floor are opposite half-spaces. the space pyramid of wall i is SP: lJ(fr. The JPs that lie entirely in this region are 1001 and 1101. The space pyramid for wall 4 is the region below the dashed line. tained in this SP are 0001.2.13 Key blocks of wall 4. The JP of any such block includes the intersection of both half-spaces of the repeated joint set and. the JPs entirely in the SP for wall 1 are the regions 0110 and 0010. Therefore. and 0011. 1101. The space pyramid for wall 3 is the side of this line that contains a horizontal vector whose direction is 298". These are 0010 and 0100. 7. as given in Table 7. For wall2. 0010. The removable blocks of the floor are those whose JP lies entirely inside SP : U(2) and therefore entirely inside the reference circle. and 1100. the JPs of the removable blocks are cousins (see the section on symmetry of block types in Chapter 4). Figure 7.). In Fig. projects along the circumference of the great circle for . which are cousins of the preceding.l2the dashed line is the projection of walls 1 and 3. From (7. Similarly. the SP for wall 3 is the region on the unshaded side of the dashed line. the space pyramid is on the ruled side of the dashed line. The joint pyramids con- 1111 / -$t *$/ Figure 7.L3 shows the stereographic projection of walls 2 and 4. therefore. Now consider removable blocks with one repeated joint set.

7. Not only can they be large. by walls 2 and 3 is given by the intersection of u(frr) andu(fir). 7 / / / / / ---t S . consider the wall/wall edges. and 3011. Applying this test to the JPs identified in Fig.1s).3. Err. on Fig. These walls are opposite walls 2 and 3. Figure 7. the Sp for edge E* formed. 1103. JP 0010 deflnes a removable block in edge Etn. so edge E. The key blocks of the roof/wall edges. Err.I4 produces a list of additional removable blocks. a block is considered removable if its JP plots entirely in the appropriate SP.Err. and Eor. this is the region on the unshaded sides of the dashed lines. the following are entirely outside the reference circle and thus determine removable blocks in the roof: 1131.14 identifies the JP codes for all segments of all the joint great circles. Similarly. Key Blocks of the Edges of the Underground Ghamber First. 1301. 7.15. 218 Block Theory for Underground Chambers Chap. From (7. All the results of this section are summarized in Table 01 13 fsotg o c) 0031 / I / / o (r) Figure 7. n is the cousin of the removable block of edge Err.14 JPs with one repeated joint set. are poten- tially dangerous. Ed. As previously. 1311.ges Erz and Ern have no removable blocks.t -- {. but they occur high in the excava- . formed by the intersection of walls 1 and 4. 1031. it is formed with JP 1101. For example. that joint set. There is only one removable block in Err.

1100. Crnr. 0010.14 1031. l3ll.13.dge Erc 0010 0030. Edge Eto None None 7. Concave Edges.3. 001 I 3001.14 0310. 1003.0130.14.14 Edge.Eee 0010 0030.12.0310. The space pyramids for the roof corners ate shown in .13.7 as a guide.13) and therefore determines removable blocks in each edge.14.15 EdEe Ezt 1tor 1131.1ll0 1103.13. 0013 7 .13. 0130 7. 0310 7. 0013.13.7. 7.0030.15 00r0 0030.15 Edge.14 Edge Ezs I101 I 131. Floor. Using Fig. The removable blocks of all edges are summarized on Table 7.3 Summary of Removable Blocks for the Example Considering Roof.14 F. 0311.14 1 301 1.0100 3100. 7.13.1301.1301.0113. 7. and Concave/Concave Corners Removable Blocks with: Position No Repeated Joints 1 Repeated Joint Reference Figure Roof (lZs) 1101.3100.14 Corner Czrs 1101 1131. 7. 1300. 3 1 10.14.0310 7. The key blocks of the floor/wall edges are found similarly.7. Since all the vectors of the space pyramids of the floor edges are directed upward. 7. the removable blocks of each of these edges have been determined. 1301.7.14 0030.0310. 7. 0310.15 Edge Ers None None 7.14 Edge E+s None 301 I 7. 3001. 3011 EdEe Erz None 3110 7. 0031. 7.15 Corner C1a6 All other corners None None tion.7. 1031 7.14 Edge. 0013. 1130. 0003. 7. Crnr.13. Walls.6to test the data on Fig.7.14.13. I 103 7. 7.7. 7.0010 3110.12.12.15. 1103.1r. These have been numbered as shown in Figs. 1301. 1 1 1 3 .0013 Wall2 (Wz) 1101.Ess 1101 1131. JP 1101 lies in the SP for both edges Ert (Fig. The most important regions are the roof cornersl C125.7 . 0300.11. 7.14 0030.1103 Wall4 (I4r+) 0001.I2) and ELs (Fig.7.3011 Floor (We) 0010. Edges E* and Errhave no removable blocks. 7.14. 1301.7.7.15 Edge Er+ None 3001 7. 0310.1101 1301.14.Er+ 0010 0013. 7. Crrr.1103 7. and. 7.0013. 1031. 0130. 113 Wall3 (LTg) 1001.12.7.14 1131.7.0013 Wall | (W) 0110.10 and 7. 1103. 1011 113I.1103 7. 7. RemoYable Blocks of the Corners There are eight corners in the gallery.Example: Key Block Analysis for an Underground Chamber 219 TABLE 7.14 Edge Ezc None 3100 7. 0030.15. using Fig. there can be no key blocks in the floor edges if gravity supplies the only force. 7.7.

2.16 shows the traces of joints in each of the walls. Fig. For the current example.9. This can be verified on the stereo- graphic projection. removable blocks of C2s. 7 1111 Joint 3 /s> €/ '% 1001 = 1011 Figure 7. 7. We will demonstrate how to delimit zones of potential key blocks ("danger zones") in the walls and in the four wall/wall edges.15 Key block of edge 84. Using the List of Removable Blocks to Locate "Danger Zones" The previous analysis determined the list of removable blocks presented in Table 7. showing the edges of joints on the walls. Since 1101 is a removable block of C4 s. Fig. 7. Therefore. The JP diagrams for the floor corners were shown in Fig. Such a map can be developed during construction as the excavation surfaces are exposed. as will be shown. Opposite corner C23.3. roof.8. Only corner Crr proves to have a removable block-determined by JP 1101. 220 Block Theory for Underground Chambers Chap. it could prov. is corner Crnu. .If a joint trace map is available. we determineiio* symmetry that 0010 will be a removable block of Crnu. Although there is only the one potential key block. the removable blocks can be delineated. and floor. have cousins in Ctnu. critical and will have to be examined further.

N N Figure ?. .10). 7.16 Geological trace map of walls (compare with Fig.

Table 7.3 and below planes 1 and 4. search of the trace map fails to turn up a polygon of this iype. it would be prudent to consider block . the search is for polygons simultaneously above planes 2 and. Because walls 1 and 3 do contain the line of intersection ofjoints 1 and 2. and 4 andabove joint 3. The removable blocks of the four walls and the four watl/wall edges (edges 1 to 4) are listed in Table 7. Edges Errand. With inclined geological map planes. Note that the stereographic projection line for walls 1 and 3 pass through the point of intersection of great circles for joints 1 and 2 (Fig. The long polygons between joints 1 and 2 in these walls belong to blocks with a long Jdge that fails to "daylight" into the excavation. The joint planes are identified by number in the geological map.3. Since 1110 is in wall 2. 0011. each joint's dip direction points into that half-plane of its trace belonging to its upper half-space. In the case of a horizontal geological map.ll20 must also be a removable block in wall2. For block 1001. There is no removable block to be found in wall 4. blocks 1011. there are no potential key blocks in the vertical edges of the chamber when it is oriented as given.E as a potential key block also.18 shows a very large key block (0011) in wall 3 after such a rotation. labeled A and -8. and a short trace of joint 2lies above the stipled region. of type 1120. and the upper half-plane of any trace belongs to the upper half- space of the respective joint. Therefore. although there is a removable block type in edge . Figure 7. is identified as a removable block. in wall 1.I2).) Figure 7. . Similarly. all the map planes are vertical walls. Block 1101 was determined to be removable in edge Err.Er 4 (0010). Similarly.4 lists the real and potential key blocks of the walls of the gallery. the stippled polygon c. For block 1101 a search must be made to locate any polygons that are simultaneously below joints 1. the removable blocks without repeated joint sets are 1101 and 1001. Consider wall 3. and 0111 were determined to be nonremovable. Ernwere determined to have no removable blocks. This more com- plicated situation was discussed in Chapter 6 and a special map code (MC) was introduced to facilitate solution. The polygons stipled with dots. of type 0110. 7 . 222 Block Theory for Underground Chambers Chap.16. are of type 1001 and are therefore removable blocks. 7.17 shows the shapis of these blocks and their positions relative to the chamber. Examination of all the polygons fails to reveal any one with this block code extending across the inter- section of walls 2 and 3. Wall2 also has a block of type 1210 (block E). either half-plane of a joint trace may correspond to the upper half-space of the joint. Z presents block D. If the chamber is rotated 90'. the long walls take up the orientations of the present short walls and some of the key blocks shown in the short walls become more important. 7 In the case of the geological map of Fig. WaIl. Since both 1100 and 1110 are removable blocks of wall 2. (There were none found in the wall/wall edges. 2. This is why the traces of joints I and 2 are parallel in the geological maps of walls 1 and 3.

17 Key bbcks actually located in the walls. 4\ \g /\ N N g. Figure 7. .

.hr N 5 Figure 7.18 Key block when the excavation is rotated through 90'.

the largest walls are W r and W 3. E Wall 3 1001 A Wall 3 1001 B Wall 4 None None Edges Erz.the volume of the larger block is then eight times the volume of the former and since both have the same sliding direction. with four times the areas of W z and W 4. can be martialed to choose an optimum orientation for instability as required for methods of mining that employ self-caving of the rock mass. Of course. The Most Gritical KeY Blocks The most critical key blocks tend to have the following attributes: 1. above polygon E. Fig. cuts off the block E from behind. For example. the net sliding force of the larger block is eight times that of the smaller. CHOICE OF DIRECTION FOR AN UNDERGROUND CHAMBER From the preceding it is apparent that the choice of direction for the long dimen- sion of an underground chamber can markedly influence the types. None None Et+. 2. and sizes of potential key blocks. In the following sections we will consider how to make a wise selection for the direction of a chamber in order to promote safety and reduce need for supports. ?. Similar arguments. Ezt. and3. Figure 7.4 Real Removable Blocks of the Underground Chamber of Fig. The most extensive joint sets are 7.Choice of Direction for an Underground Chamber 225 TABLE 7. Their space pyramids contain steep vectors. They involve joints of large extent. 2.19 shows two similar blocks with edge lengths in the proportion 2: l. 7.20 presents the JPs of these three joint sets alone as well as the projection of walls . They belong to the largest free planes. the maximum key-block dimensions allowed by wall dimensions can only be realized if the joints have sufficient extent. Et+ aPotential key block if joint 2. In the previous example.16 Location Name on the Map Wall 1 0110 C Wall 2 tt20 D Wall 2 1110. numbers. The size of the free surface provides an upper bound on the dimensions of removable blocks and the dimensions dictate the maximum volumes. in reverse. 3.

7 Volume = 8 V Figure 7.2. . Block Theory for Underground Chambers Chap. / / / / / / -'/\ / / 0002 /' 0012 Figure 7.20 JPs of joint sets I .19 Relationship between relative volumes and relative edge lengths. and 3 only.

If JP belongs to a remavable block of Eti. Edges' and Corners Since geometrical relationships connect corners. If we assume that JP is removable in Eii... then JP c SP(E i) Introducing (7. is a free plane i i : 1. Notation: W. edges.3e) Proposition 2. This section explores such relationships. edges. edges..37) can be written as SP(Crie) : SP(E i) n SP(E &) n SP(E &) (7. We consider here only concave rack edges and corners. and i : 6 identifies the floor. W1. the most critical key blocks. JPc(SP(et")nSP(IZ1)) JP c SP(H/. Indeed. C. SP (l/r) is the space pyramid of W. 6 SP (E i) is the space pyramid of edge E" and SP (Crir) is the space pyramid of corner Crro In a concave chamber. If JP belongs to o removable block of C. then JP belongs to a removable block of Wt and W.36). and W* The order of indexes has no significance. E.37) As a consequence of (7. in the interior of a prism. and planes of a prism. the space is convex and its edges and corners are accordingly the intersections of half-spaces as follows: SP(E r): SP(P|") n SP(fj) (7. convex edges and corners are discussed. then JP belongs to a removable block of W. and Wo. Relationships between Key Blocks of of Direction for an Underground Chamber 227 1 and 3. In this way.) and JP c SP(|4li) (7. JP c SP(Crjk) .to is the corner formed by the intersection of W.36) to (7. i :5 identifies the roof.38) may be considered to constitute a theorem of linkage in underground chambers. is the edge formed by the intersection of planes i and j. for i : l.. Since JP belongs to a removable block of C.4 arc walls. and corners.36) SP(Gio) : SP(I4IJ n SP(I4lj) n SP(14/e) (7. and planes of a prismatic underground chamber. Proposition 1. the direction of these walls was chosen precisely to contain the line of intersection of joint sets 1 and2. In the next section. equation (7.38) Equation (7.36). They together establish three propositions linking removable blocks of walls.1p. Wr. of the largest walls have been avoided. It can be seen that none of the eight JPs are removable in either wall L or wall 3. there must also be relationships connecting removable blocks of corners. concerning intersections of chambers.

4t) Figure 7. and corners.1p. w6 (f loor) I I I + atrq.e. then JP belongs to a removable block of edges E. edges. JP c (SP(E i) n SP(Eje) n Sr1r.37). E*. n sr1nzrll SP(w) so JP c SP(ry.) and JP c SP(ft r) and JP c SP(14l*) (7. ./t IlTCr2s-Ers+Crasll | r\ _i /1 I | w6-l (ttoir)- rrr .i.38). wo "T" |ll*| | -.*+. Again. JP c (SP(W) .[l.r Figure 7. If JP belongs to a removable block of C.+-E16-+C1a6 +\ -rz-.".---34 Czro+Ese+Cga6 \t I + I I I I I r. since JP belongs to a removable block of C..2I diagrams the geometric connections between all walls.. then introducing (7. 7 Introducing (7..40) Proposition 3.Tir)* I l' i" (froorr | rll I i/'"" I + rt l-zgs-Ess+Cs+s I r/ | t -zJ. JP c Sp(Crr*)._'tl-Ll4 { . edges. and JP c SP(Eio) and JP c SP(E o) Q. and E.-"3.J lll r +ll l-llll eju**o*eou+- ti' Block Theory for Underground Chambers Chap. and therefore JP c SP(E.21 Linkage diagram for walls.

and E26. if a JP is not removable in wall 4.10. it is not removable in any edge or corner contiguous to that wall. and C1+o'If a JP is not removable in E 3 4.Choice of Direction for an Underground Chamber and corners. since line I contains /1 . Table 7. if a JP is not removable in a wall. a Jp that is removable in a corner is removable in both the edges and walls that are contiguous to that corner. In the present example. JPs 1000 and 1011 cannot determine removable blocks in the right wall. if it is not removable in an edge. Erz. E+e. Draw a line through the two points of intersection of a pair of joint circles : for example. the formulas in this table were derived assuming that the free plane does not contain any line of intersection. there are three removable blocks in each wall. In other words. it is also not removable in C3 n. as in Fi5. Draw the stereographic projection of all sets of joints. Line I is the direction of a vertical wall that parallels the line of inter- section of joint sets I and 2. For example. The relationships described above make the work easier. nr. radiating from the center of the reference circle. it is also not removabl e in E1+.2l. However. a JP that is removable in an edge is removable in both walls contiguous to that edge. For if you can choose the direction of W. and En. generate all such lines on the stereographic projection. nu. drawn for the intersection of joints I and Z.. Wr. In the example considered. all contiguous edges and corners can be found. (In an inclined chamber. with four joint sets. the removable blocks of each are symmetric cousins. Using Fig. there are six such lines. Determine the removable blocks of the right and left walls for all i.7. it is not removable in either corner attached to that edge. According to proposition 1. the lines are replaced by great circles that radiate from the normal to the roof and floor.Crnu.7. We then define the right wall of line i as the wall parallel to line i and with the rock on the side denoted as "right. Repeating for every combination of joint planes.) 3. the JP cannot be part of a removable block of either the large free sur- faces of the corners or long edges of the excavation' Procedure for Choosing the Direction of an Underground Ghamber 1.22. to avoid a particular JP. Note that since the right and left walls determine symmetrically opposite half-spaces. According to Table 6. For example. Ern. and Wu as well as in edges Eru.5 presents these results for the example (r : 1 to 6). line 1 of Fig. there is one less removable block in each wall parallel to lines 1 to 6 precisely because these walls do contain the line of intersec- tion of two joints. Conversely." 4. Arbitrarily denote right and left sides of each line. And by propositions 2 and 3. then all the edges and corners of these walls are also safe from key blocks of that JP. and W. 7. and it is also not removable in C. For example. and JPs 0111 and .Crnr. and C . a JP that is found to be removable in Clru must then be removabl e in Wr. none of which is repeated.22- 2.

4 t47.?2 Analysis of the effect of direction on the key blocks of a prismatic underground chamber.4 r77. ooot"" '10111 . .O 0110 1001 0100 10il 230 .5 Removable Blocks of All Walls Containing the Line of Intersection of Two Sets of Joints Direction Removable Blocks of the Line Joint Intersection Right Left Number Sets (dee) Wall Wall I lr2 28.2 1110 0001 1100 00il 5 2.1 1101 0010 1000 0111 3 1.8 1110 0001 0100 1011 6 3.\ 1001 1oo0 \i 1. 1111 4et 1/ Joint 2 \ obhr // Joint 4 / .7t /'.*.4 73.3 133.').4 1001 0110 1101 0010 2 lr 3 58.' oorl s i Figwe 7. TABLE 7.' oom t /" '+Hrr.0 1101 0010 1100 0011 4 2.

2 1110 0001 I 100 001 I 1101 0010 415 t33.0 133. as given. .O 28.8 177. steep key block in an excavation should be avoided at all costs. all these directions of walls have three removable blocks each.4 1001 0lI0 1101 0010 1011 0100 1.2 147.2 28. The pos- sibility of having such a large.4 58. Determine the removable blocks of the right and left walls directed in each of the angles bounded by lines i (i : 1.1 73.O I101 0010 1100 0011 1000 0ll1 3.6 Removable Blocks of Walls in the Angle between the Intersections of Two Sets of Joints Line Start End Removable Blocks Numbers Direction Direction Bounding of Angle of Angle Right Left Directions (dee) (dee) Wall Wall 617 357.4 73. The removable blocks of the right and left walls are symmetric cousins.1 1101 0010 1000 011r 1001 0110 213 58. Figure 7.O 0110 1001 0100 10ll 1110 0001 The apptication of this procedure can be demonstrated by examining the consequences of a rotation of the long walls of the example considered pre- viously.Choice of Direction for an Underground Chamber 0100 cannot determine removable blocks in the left wall. the direction of the long wall will belong to the angle between line 4 and line 5.If the chamber is turned through 90o. 7. The results are listed in Table 7. With walls I and 3 directed to the N 28" E.18 shows then the maxi- mum key block of JP 0011 within the long left wall of the chamber. or the other of these JPs would be added to the list of removable blocks.17 . Since none of these walls contains the line of intersection of any pair of joints. (However.) 5. interchanging the directions of the long and short walls. even then the corresponding blocks would be only narrow slivers in the wall.6. the key blocks are those sketched in Fig. TABLE 7. 6 in the example). If the direction of the wall were rotated a small amount in either direction from that of line one 1.8 1110 0001 0100 1011 1100 001 I 516 147.

b.23). midroom. and roof heights and the various edges and corners are identified.23. For example.x $ 4.232 Block Theory for Underground Chambers Chap. In building trades. We have introduced the following notation: .f7 $.23 SP and EP for the inside edges of intersecting chambers..25 show horizontal sections through intersections of chambers at floor. E\2 is the edge of walls 1 and 2 where a chamber parallel to wall 1 intersects a chamber parallel to wall2 (Fig. ../+p Figurc 7.7. and 7.24. these are known as 'ooutside corners" as opposed to "inside corners" of conventional rooms. 7. Figures 7. E'u denotes the convex edge of an exterior corner where walls i and j intersect. 7 I]UTERSECTIOIIIS OF UIUDERGROUIIID CHAMBERS The crossing of two underground galleries creates convex edges and convex/ concave corners where these edges meet the roof and the floor.

23. that points into the free space of the intersection. -t-- / /\.t meets the roof or floor (wall k. 7. denotes the exterior corner where convex edge E'. .43) .t Figure 7. In every case.42) sP: -EP:U(fr. as shown in Fig. respectively). is the normal unit vector of W. Remov- able blocks of a convex edge therefore have excavation pyramid EP given by EP:L(fr. r0. CrtJr.)^L(fi)r) (7.Intersections of Underground Chambers ''':A.)uU1fi) (7.4 SP and EP for the walliwall/roof corners of intersecting chambers. with k : 5 or 6. Wt denotes the four walls involved in the intersection. . Convex Edges of Ghamber Intersections The rock mass delimited by the convex edge E"" is L(fi.) ) L(tfi ).

The criteria for removability of a block in the edge E'. the removable blocks of edge E\.L4 and 7.15. Since the four convex edges of the intersection have a larger SP than that of a wall.o (7.t are therefore stated by JP* A (7.7. 1110. For example.15) and for JPs with one repeated joint set (Fig. are those corresponding to JP's 001Q 0110. the removable blocks of all the edges were determined for the wall directions given previously in Table 7.25 SP and EP for the wallywallTfloor corners of intersecting chambers.23 shows the EP and SP for each of the convex edges of the intersection.14).ft)) or JP+ @ and JP n (L(fr. Results for all the edges of the intersection are given in Table 7.) n r(fr)) . 1100.44) and JPc (U(ft)vU(. 7.xt: Figure 7. 7 lilll EP '\ cioe / >t'. more blocks are removable in the edges of intersections than in the . both for JPs lacking any repeated joints (Fig. 7.234 Block Theory for Underground Chambers Chap. 7.0100.4s) Figure 7. and 1101.1. Combining this with Figs.

can be divided into two blocks BPt and BPr. Corners of Ghamber Intersections The external corners Ctt. discussed in the preced- ing paragraph. However. are JP+ g and EPnJP:@ or JP* @ and JP c SP where EP : L(fr) u (L(fi.N 0110 3110 0311 3001 0111 3111 0030 30ll 0010 0130 0013 00ll 0113 0003 0001 0310 0031 E\z 0010 il01 0030 3100 1301 0110 0310 0113 I 131 0100 0013 3110 1103 1ll0 0300 ll13 1300 I 100 0130 1130 E'zt 1001 3001 1300 3110 1000 3000 1131 3100 ll0l 1031 1103 1100 1003 1113 1110 1301 1130 walls of a chamber.lntersections of Underground Chambers TABLE 7.r1. Such edges may proye to be critical locations in an under- ground chamber. the SP is not convex and unlike the edges of an intersection.) t (7.48) ^fr)) . Unlike prismatic chambers. each block of the external cor- ners C'a1. using Shi's theorem for nonconvex blocks. where BP1 : L(fro) n JP (7'46) and BP2 : L(fr A L(fr.k where one chamber intersects another are com- plexly shaped.7 Removable Blocks in Convex Edges of the Intersection of Two Underground Chambers Removable Blocks with: Position 0 Repeated Sets I Repeated Set E's+ 1101 0010 1131 3011 0310 1001 1301 1003 0030 10ll 1103 3001 0013 0001 13lr 0003 03tl 0011 1031 0031 E. the EP is not convex. ) n JP (7 . or concave surface excavations.47) ') Then. the criteria for removability of a block from corner C'..

236 Block Theory for Underground Ghambers Chap. The results are listed in Table 7. 7. and C'1a.) n Z(fr)) or SP : U(fr) n (U(fij) u a(ti.25.49) The SPs and EPs of CI js and C'. parallel chambers may be called a wall pillar or a rib pillar. Note that the results divide into symmetric pairs: E'r.16 are shown in Figs.I4.8 Removable Blocks in Exterior Corners of the Intersection of Two Underground Chambers Removable Blocks with: Position 0 Repeated Sets I Repeated Set C'st. we can determine the removable blocks cor- responding to JPs with no repeated joint sets. 7. The key blocks of all these pairs are cousins. Etini C\zs and. sP: _EP : U(fr)n *(L(fi.)) (7. and. 7.8. and E|ni E'r. respectively. and C'rro. TABLE 7. 7 From (7.45) and (7.24 and 7 .L5. And with the use of Fig. 7. An example is shown in Fig. the resulting rectangular or square column pillars are all that is left to support the .46). This is a critical part of any under- ground complex because its failure might bring down the roofs of the adjacent openings.1. we can determine the removable blocks of JPs having one repeated joint set. If a rib pillar is cut through on a regular pattern of cross cuts. Using these figures with Fig. 1101 1131 1311 l0l l 1103 1031 1301 3011 C\+s 301 1 C\zs 1101 I 131 r 301 1103 C'zss 1101 1031 1301 1311 1103 1131 C'sqa 0010 0030 0310 0013 C\+e 0010 0130 0310 0300 0013 0030 C\ze 0010 0030 0300 0100 0013 0r30 0310 3100 C'zta 3100 PILLARS BETWEEN UNDERGROUND CHAMBERS The rock separating long. C'3aui C'rns and C'r2u.

26 Key blocks of a wall ("rib"). Assuming that the stress concentrations in the pillars are insufficient to fail the rock itself. and their cousins. All the JP regions not touching the dashed line therefore determine blocks that are removable in the pillar.26 shows a pillar between parallel walls of neighboring chambers. Figure 7. The criteria for a key block are.Pillars Between Underground Chambers roof. Let ril be the normal unit vector pointing into one of the two openings. they are 1101. but these can be represented by a series of m tangent planes. Key blocks of pillars can be discussed using block theory." 1000 : i--f mm 001 Figure 7. Let the unit normal vector of tangent plane W.. then the EP for the rock of the pillar is determined by EP: U(fr)nL(fi)) (7.27 shows several pillar cross sections. Wafl pillars. tP+@ and JPnEP:@ The projection of the EP for the pillar is the dashed line of Figure 7. For this example.26 and the SP is all the space not touching the line. 1110. Figwe 7. An actual pillar will tend to have curved faces. Column pillars. as . 1100. a pillar is still vulnerable to progr€ssive failure following loss of key blocks.s0) 1t11 .

EP is then a line parallel to the axis of the pillar. The EP for the pillar is then given by the inter- section of all L(fi.2il Key blocks of a pillar. All nonempty UPs therefore define removable blocks except the two that contain the axis of the pillar and its opposite (Fig.s1) For a pillar of constant cross section. Not only is it easy to judge whether or not EP and JP have common points (or if JP is entirely contained in SP). The latter is more difficult to appreciate using vector methods. COMPARISON OF VEGTOR AITIALYSIS AIUD STEREOGRAPHIC PROJECTION METHODS Since the stereographic projection shows both EP and SP for an excavation.27). Block Theory for Underground Chambers Chap. but also whether the boundaries of JP and SP are close. In the case of a vertical pillar EP:12 (7..): EP : n (r(ltJ) m J= I (7. 7. 7 #ffiffi RigureT. all the block problems of this chapter can be performed on the stereographic projection.s2) A JP will define a removable block in the pillar only if its intersection with EP is empty. tD. pointed into the free space. .

edges and corners of an opening without intersections or reentrants. To determine whether or not two convex pyramids intersect. A calculation must also be made to determine if the second contains any edges of the first.Comparison of Vector Analysis and Stereographic Projection Methods 239 For the calculation of key blocks of walls.47) the criteria for removability of a block may be written (L(ft. It is sufficient to compute the edges of JP and determine if each edge belongs to each of the half-spaces of SP. Although all these problems can be solved with vector equations. whether or not JP is entirely contained in SP. It is then straightforward to determine. in the case of complex blocks of intersections it seems preferable to use the computer to produce a stereographic projection which is then examined interactively. the SP is not convex and vector methods are not con- veniently expressed in terms of SP. by vector methods. it is not sufficient to determine if the first contains any edges of the second.)^L(tD)) nJP: @ (7. . the EP criteria are preferred.s4) Each of the equations above represents the intersection of two convex pyramids. From (7.s3) and L(fi) nJP. @ (7. For the complex blocks discussed in connection with intersections of underground chambers. Instead. SP is convex for all elements of the excava- tion.

Temporary tunnels for mining are usually left unlined. They have been dug for access. chapter 8 Block Theor>l for Tunnols and Shafts Tunnels are among civilization's oldest achievements. and under ridges to avoid extensive surface cuts. the stability analysis using block 240 . Tunnel direction is usually determined before breaking ground. However. whereas natural hillsides undercut by surface excavations often have convex slopes. and mining. and the direction selected for a tunnel greatly affects its excavation and support costs. but because the span of the tunnels is far smaller than the dip length of the slope. meaning that much smaller key blocks are involved. transportation. whereas convex excavation surfaces have large space pyramids and many key blocks. In previous chapters we observed that concave excavation surfaces have small space pyramids and few key blocks. Such tunnels tend to be safer than their cut-slope alternatives not only because the weak rock of the weathered zone is avoided. shelter. tunnel engineering demands on-site decisions. We place highway and railway tunnels under a sidehill to straighten and shorten a route. early tunnels were probably completed with little or no engineering calculations. defense. In this chapter we examine exactly how block theory can be applied to choosing an optimum direction. water supply. because the main body of rock to be pene- trated remains essentially hidden until actually encountered in the tunnel face. Even now. tunnels have concave slopes virtually everywhere. drainage. Moreover. When there is no lining. as we have seen. So are long-life tunnels for water power and water supply in good rock and for transportation in very good rock. Except for surveying. mountains of rock are directional with respect to excavation.

in all tunnels the state of stress acting around the tunnel can influence the stability of key blocks. Horizontal and vertical tunnels are relatively easier to excavate and to line. hollow solid of constant cross section. The system of blocks can then be analyzed to determine the possible loads and load distributions on the lining under dynamic forces for example by an explosion or earthquake. hydraulic. inclining a tunnel is not compatible with its purpose and a horizontal cylinder is required. enlargements. Calculations of the stability of key blocks under gravity. What is special about tunnels in this regard is their curved surfaces.1 sketches a tunnel under construction for a hydroelectric power project. inertia. Figure 8. The main part of the tunnel is the tunnel cylinder. Therefore.Geometric Properties of Tunnels 241 theory pertains to the entire life of the tunnel. block theory calculations then pertain to the period until the lining is constructed. but the greater design freedom to choose both a tunnel direction and inclination enables a better choice of orientation with respect to key blocks. But the portals are also very important since a difficult portaling condi- tion can delay and complicate a project. Also. key-block problems tend to be less severe than in the tunnel cylinder or the portal. Portals are generally more difficult than the tunnel cylinder. particularly in water power projects. The working face of the tunnel is another element. We use the term tunnel to describe the whole system by which passage is obtained through a rock mass. and bends of tunnels are not discussed in this chapter. *By "cylinder" we mean a long. GEOMETRIC PROPERTIES OF TUNNETS Tunnel directions. Then the methods presented in Chapter'l can be applied.* However. which generate blocks with curved faces. Since stresses are concentrated here. with low friction angles along discontinuities. Such cases can be analyzed approxi- mately by replacing the tunnel section by a series of tangent planes. however. The horizontal tunnel will emerge as a special case of the general theory. and structural loads are given in Chapter 9. the general case of inclined tunnel systems will be treated. The commonest system includes two portals and a horizontal cylinder. Often. Not only is weathered rock encountered. for stability or hydraulic smootheness. . The shape of the section can conform to any continuous locus. In this chapter we continue the analysis of the geometric conditions lbr the formation of remov- able blocks. The inertia force of the key block adds to the sliding force and the block presses against the lining. When a lining is built. The complex cases presented by intersections. Elements of tunnels. Water forces on key blocks have to be considered for pressured water tunnels. and the excavation surface is concave. but the excavation surface at the portal has a larger space pyramid than that of the tunnel cylinder. vertical and inclined cylinders are also used.

3. The upper part is a smooth curve lacking angles. closed curves.4(b). But excavation and lining may be more difficult and more costly.6 show various shapes for the tunnel cylinder. 8 I ntersection Main tdnnel cylinder Figure 8.2 shows smooth. 8. The shapes in Figs. .5(c) represent hybrid solutions. Tunnel shapes. Figure 8.1 Geometric elements of a tunnel. Such shapes reduce stress concentration and diminish the number of key blocks. Block Theory for Tunnels and Shafts Chap. without any angles. Figures 8. to minimize stress concentration and reduce the {b) (a) Figure 8.2 to 8. 8. and 8.2 Continuously curved tunnel shapes.

5 Polygonal and horseshoe shapes.Geometric Propenies of Tunnels (a) (b) Figure 8. .3 Tunnel cylinder shapes with straight sides. (b) Figure 8. Figure 8.4 Additional shapes for the tunnel cylinder.

2U Block Theory for Tunnels and Shafts Chap. We now introduce coordinates f o.a(a) and (c). I (a) {bt Figure 8. and 2o in directions attached to the tunnel. for easier construction. occasionally found in transition sections. Although seldom designed. it will be helpful to introduce a tunnel coordinate system. In particular. jo. vertical for a shaft. 8. Figures 8. This is evidenced by examination of natural tunnel sections and can be shown readily using key-block theory.5(a) and (b). could result from loss of a key block in an initially flat roof. The lower part is produced by straight-line segments. Goordinate Systems Let the axis of the tunnel cylinder be denoted by unit vector 6. The resulting asymmetric shape with a sharp corner provides a high stress concentration which may stabilize other blocks riding behind. The global coordinate system we have been using has i horizontal directed to the east. ! horizontal to the north.horizontal for a horizontal tunnel. and 8. its intersection with a system of joints produces blocks with a curved face. In order to apply block theory to this case.6 show polygon shapes. number of key blocks.6 Vaulted tunnel sections. Ideal polygon shapes are also created by partial collapse of sections of a tunnel as blocks drop from their initial positions. for example. and so on. . BTOGKS WITH CURVED FACES Since the surface of a tunnel excavation is cylindrical. asymmetric shapes can pro- vide greater stability than symmetric shapes.6(b). and f directed upward. 8. The vaulted roof of Fig.

1) then give /t'\ l.7. 8.?o is its strike. From Q.1) Yo -a Xxo:d1x.'(2xd) Using the dot product of each pair of *o.7 Tunnel coordinate system.9o.90.5000) Equations (8. o -0.1) are right-handed.and. we have (*o x 9). lished that these coordinate vectors are mutually orthogonal.20 defined by (8.20 for the example are plotted on a lower- focal-point stereographic projection in Fig.| 0 l: I lo.0.2o in turn. while .f o is directed up the trace of the dip vector of this plane.8660 0.5000/ The coordinate vectors fro. assume that the plane normal to the tunnel axis has dip d : 60o and dip direction F : 0".io. It will now be established that the coordinate vectors *0.0.iofo. The plane pe{pen- dicular to the tunnel axis is the plane of .Blocks With Curved Faces 245 zo xo xd (8. it is easily estab.9660.r*) \20l \ 0 0. . Using these relationships. directed according to 2 x d.s000 l. the coordinates of the vector d normal to this plane are d: (0. As an example.@) N A az Yo ---+ Figure 8.2o :(t x d) x @ x (2 x il)..7).

Consider one such tangent to the tunnel at point Q.246 Block Theory for Tunnels and Shafts Chap.Qa x Q x fi) :(axt)x@).(AxilxG)) :KAxflxAp>0 Therefore.B. C:-AxC. Clockwise rotation of (.andwhose edge makes an angle 0 with jo (measured positive in a clockwise direction) as shown in Fig. 8. *0. Local Goordinate Systems for Points on the Tunnel Cylinder Planes tangent to the tunnel cross section are all seen as edges in the fto9o plane.t).f o) in the plane 2o: 0 by angle 0 about the origin to{ t0 90..E Position angles and local coordinate systems for points on the tunnel rylinder.8.20 comprise a right-handed.90. t{o} (c) {d} Figure 8. orthogonal system. 8 ThensinceAxB.'.f. (*o x 9)-20 : -(2 x il x (. .

The angle 0 is always measured clockwise from io to i(0) or from . In Fig... Then for any angle 0 between 0 and 360' we can locate one point Q. : U(fr(0. ('.Blocks With Curved Faces generates new vectors i(0).)) n JP. t(0) is the tangent vector of the tunnel surface at point Q.@. . we will adopt the convention that the vector 2o : d is pointingfrom the tunnel section toward the observer. -. i : 1. fr(0) points out of the tunnel. a continuous closed curve. In all views and discussions of this vector.2) Suppose that the shape of the tunnel section is convex (i.)) (8. Let B.: . fr(0). so that the tunnel curve is determined by the locus Q@)B. while ffi .9.2). n(q is the normal vector to the tangent plane at point Q. into the rock.3) The excavation pyramid for the complex block is then EP : i=1 0 U@.4) and a block is removable if JP + g and JP n EP : @ . The latter begins to be a close approximation to the smooth curve of the theoretical tunnel section.:1. 8.9(a) the tunnel curve is unsatisfactorily replaced by two tangents (i.8. seen on edge in Fig. 3. without reentrants).xi.) (8. to the tunnel wall at point p. First choose m points along the curved boundary: Q(?t).9(b). 8. respectively. The curved tunnel surface is then replaced by a multiplanar locus.4 is shown in Fig. ot a straight segment on the tunnel boundary such that: l. The angle d is the tunnel position angle of point Q@. be the point of intersection of tangents through Q@) and Q(0.9(c). .90 to fi(0). the boundary of the tunnel cylinder will be approximated by m tangent planes.f o). as shown in Fig. .. Q(0^) and construct a tangent line through each. 8.Q(0*) A complex rock block between Q(0r) and Q(0^) is the union of rz convex rock blocks with the block pyramids BP.*).. 2. . . as shown in Fig.e.. The projection plane in the projections of curved blocks will always be the plane of the tunnel cross section (fo. m (8. Q@).Q(0r)BrQ@r) .): (:"i . 8.9. The case for m:3 is in Fig.. .e. tangent and normal. B^_. 8. Excavation Pyramids of Gurved Blocks To analyze the curved blocks of tunnels. 8.. .. m .

248 Block Theory for Tunnels and Shafts Chap. add 360' to all 0r having index i + 1 or greater.5) is true even if the number rn is so large that the tangent segments . values increase uniformly with increasing t.4).0. (If angle 0 crosses : 0 0 between 0. is a small. : + the whole space and EP n JP JP @ and therefore JP cannot satisfy equation (8. With Q@r) established at the extreme counterclockwise end of the curved segment. . From Fig. Thus such blocks are not removable. by tangent planes.0. and 0. 8 {a} n(02) (c) (d) Figure E. positive angle.*. EP given bv (8. then - EP : u U@@)) : U@(0.5) Equation (8. 8.a) encompasses .) Then for all i. all 0. If 0^ 0t is greater than or equal to 180".) u U(n(0^)) (8.*r.9(d) we can see that if 0^ 01 is less than 180'. between given points.9 Representation of the tunnel cylinder.

Then. The numbers of all combinations of half-spaces are as discussed in chapters 5 and 6 (Tables 5.(n' . The respective formulas have been transferred to columns 1 and 2 of Table 8. JP is a removable block of a tunnel if and only if +A + JP. This is because the tunnel axis belongs to each of the half-spaces whose union determines EP and therefore EP n JP must contain d and accordingly is not empty' Every JP that does not contain the tunnel axis. Proposition. Also.1 and 6. the numbers of nonempty joint pyramids are the same as discussed in Chapter 5. TYPES OF BTOCKS IN TUNNELS In this section we discuss the numbers of blocks of different types that are created by the intersection of a tunnel cylinder and a rock mass with n sets of joints. We will demonstrate how to locate this block later. Accordingly.Figure 8. The criteria of removability for blocks intersecting a curved tunnel surface are JP+@ and EPNJP:Q where EP: u(n(0))u u@(0^)) and 0^.10).1 to support the discussion here. from the first column of Table 8. . then.1-. This is stated here and the proof is presented in an appendix at the end of this chapter.n + 2.01 < 180' TUNNEL AXIS THEOREM A theorem concerning the relationship between the axis of the tunnel cylinder and the JPs of removable blocks will prove useful. This has been pro- duced in column 3 of Table 8.n * 2). the number of tapered blocks is 2' . the number of tapered blocks is the number of all combinations of half-spaces less the number of nonempty joint pyramids. consider a case with no repeated joint sets.1"0 shows an example of a tapered block intersecting a tunnel. the number of all combinations of joint half-spaces is 2" and the number of nonempty joint pyramids is n2 .1. If the tunnel axis d is an element of JP. As previously. has a corresponding removable block in the tunnel. This per- mits statement of the following important proposition for curved blocks. the theorem states that the JP does not belong to any removable block of the tunnel.Types of Blocks in Tunnels 249 almost perfectly match the convex curve of the excavation boundary. As an example. Theorem.

1 Numbers of Blocks of Different Types in Tunnels Number of AllNumber of Non.l) 0 n)3 ra selected repeated sets 2r-^ 0 2r-^ o 0 n) m)3 (m>3) Any m(m> 3) Cf.l) n(n-1X2"-3-t) n(n . r) 0 n)2 Any 1 repeated set n2"-r 2n(n .l) 0 n)2 2 selected repeated sets 2"-2 T-2 -z 2 0 n)3 Any 2 repeated sets n(n ... Number of Number of Number of Number of Combinations ofempty Joint Tapered Removable Infinite Repeated Sets Half-Spaces Pyramids Blocks Blocks Blocks Condition 0 repeated set 2" nz-n*2 2n-(n2-n*2) n2-n 2 n2l 1 selected repeated set 2n-r 2(n .2-^ 0 0 n2m)3 repeated sets .N gr o TABLE 8.l) . n(2n-t -2@-1)) 2n(n .2"-^ 0 C!.l)2"-z n(n .r) 2'-r -2@-l) 2(n .

a block is not removable if the extreme limits of its intersection with the tunnel surface include an angle 0^ . there is a maximum size beyond which *maximum blocks are no longer removable. The complete set of formulas are given in column 4 of Table 8. its symmetric cousin will contain -d. Thetefore. In this section we ignore the statistical question and . this means that there are nz .1. This limiting size will be termed the key block.n removable blocks. The probability of this happening is very small. The criteria for a block to be removable inside the tunnel are that JP + @ and that neither d nor -d is contained in JP. JP belongs to an infinite block of a tunnel if and only if d e JP or -d JP e JP. If ri falls exactly on the line of intersection of two planes. In the event that dfalls exactly on the line between Jps. On the other hand. there is one and only one that contains d.10 Tapered block intersecting a tunnel. The number of removable blocks is the number of nonempty joint pyramids minus the number of infinite blocks. Further. The number of removable blocks of a tunnel. 2 180'. the friction angle mobilized on sliding surfaces generally varies inversely with the dimension of the sliding surface. it would define an infinite block of a JP with two repeated joints." The largest removable block of a given key-block type is the most critical because it requires the largest supporting force.0. and therefore the number of infinite blocks with one or more repeated joint sets is shown in Table 8. THE MAXIMUM KEY BLOCK According to the proposition (8. there are exactly two infinite blocks.1 as equal to zero. the statistical distribution of joint extents dictates that the probability of actually encountering a block becomes smaller as the size becomes larger. For a given block type. From the tunnel axis theorem. Since d|s apoint inthe stereographic projection. The number of infinite btocks of a tunnel.The Maximum Key Block 251 Figure 8. For no repeated joint sets.6) for removability of blocks intersecting a curved tunnel surface. The angu- lar interval over which a block intersects a tunnel depends on both the shape and size of the block. it would define an infinite block of a JP having one repeated joint set.

enclose an angle (n.) is less than 1g0".8) (i. . r! as shown.Ut were less than 180". there is a vector r0 : rrr such that U(mo) n JP: @ [see equation (38) in the appendix to this chapter]. proofs of the propositions are contained in an appendix to this chapter.. . This could not be so unless the angle 4t .?. .B is a removable block of Jp in a tunnel cylinder.if. Then if block . the projection of such a JP is always such that (tt. establishes the following proposition : .eeD)} (8. 2. (8. There is a removable block. meaning that B c. followed by applications.e.8) ^ The maximum removable area is the intersection of the tunnel section and the space outside the tunnel with the right side of (8. * rnz fi(qr) Figure 8. 252 Btock Theory for Tunnels and Shafts Chap. i'. -B belongs to the "maximum removable area" of Jp. This block is the maximum key block.4) where 41 2 ur. The outward normals to the extreme edge projections. .. I.Ir. {u(-fi(n).eQt)) uen@). 8. and their orthographic projections in the section are r". Figure g. of JP such that the projection of Bin the section perpendicular to d is exactly equal to the projection of the maxi- mum removable area of JP. . lL. g always assume that the maximum key block is also the most critical key block. Its edges are I. Let the equation {U(fr.... .11 Projection of a Jp in the tunnel section. B. It is established in the appendix to this chapter that if Id # Jp. Q)} define the half-space containing fi whose boundary passes through Q. A formal theory will be presented first. @ if and only if 41 10 1q. Proposition on Angular Relationships for an Empty Intersection In Fig. and". therefore.L1 a JP is projected into the plane perpendicular to d (he tunnel section).7) Theorem on the Maximum Removable Area of a Tunnel section 1. . the term inside the [ ]).11 U(ft(0)) n JP ..

e. The directions of I'randfi (the orthographic projections of the extreme edges of a JP.a(a) and 8. we will show how to determine angular intervals and maximum removable areas for all key blocks of a tunnel.Computation of the Maximum Removable Area Using Vector Analysis Figure 8. The theory presented in this section can be applied usingvectorcalculations or stereographic projections.. we again use the sets ofjoints introduced in Chapter'7. Asingle JP yields a single angular interval regardless of the shape of the tunnel (assuming the tunnel cylinder is convex).3 to 8. 8. and 4.l. .12 Maximum removable area. In fact.12 shows a section perpendicular to d. it is possible to choose a tunnel shape for which the maximum removable area vanishes. the maximum remov- able area does depend on the shape. Both methods will be developed in the following sections. as shown in Figs. (i. 10" < 4). However. COMPUTATION OF THE MAXIMUM REMOVABLE AREA USING VECTOR ANALYSIS Using vector operations. In the example to be studied. at Q@ ) and Ii at Q@) may contain a removable block only if 0t and g2lie between r7. as given in Table 8.B is a removable block. o(4r) o(01) o(e2l Figure 8. enclosed between lines parallel tol'r. A similar area. The maxi- mum removable area is determined by the region that is between these tangents and outside the tunnel. Figures 8.4r 10.6. .ana 1! .2.6(b) dre examples. It can grow larger until its extreme edges reach almost t9 i'r. as seen in the tunnel section) are tangent to the tunnel section at points Q(rt) and Q(4).

42os 0.3 (see also Table 6.9042 0. Dip Direction.ztsr o. x fi). g TABLE 8.2746 _0. a. l-. fi (dee) (dee) I 7l 163 2 68 243 3 45 280 4 13 343 Matrix for Testing the Edges of a Jp The upward normal vector to set i is fi.fik\ (8. cos Br.6).| -o. The matrix of values of It/ TABLE 8.sll.10) The results are given in Table 8. given by : (sin dr sin Bs.9) lfrr\ | 0.l The next step is to compute all values of I'l : sign [(fi.3. -i\ L rFYl (1):l_l . Let (/) be the 6 x 4 matrix of values in Table 8.azor -o. 254 Block Theory for Tunnels and Shafts Chap.) fi.. _?l [-r .oslo o. i Dip. n&] 1200_1_1 13010_1 140110 23-100_1 24-l 010 34-1-100 was used in Chapter 6 to test the finiteness of a block. (g.3 Values of If : Sign [(n1 X ry) .325s\ lo. l.o t ol \-1 -t o gl .3246l lo. Hence it will be used to compute the edges of a JP.l.zo7l I \nol \-o. sin d.oosz o.l -f -o.t2z7 o.2 Joint Set. cos c.

determine a diagonal matrix (D) whose diagonal terms correspond to a signed block code D" [as described in (6.'i '(8'13) [Each row of (T) corresponds to a Ttr of (5. or if all of the elements of the ith row are equal to zeto (e).by Dn: (at a2 a3 ar) For a selected D". Each row of (1) corresponds to the intersection fr1 of a particular pair (i andi) of joint planes. I(a') 0 0 0 0 I(a") 0 0 (D): (8. (1) has m rows and n columns with m : C3 : n(n .12) [+t ira.15).: s [+t l-l if a.:3 Corresponding to (6. D".: if (r) (8'15) +1 each element of the rth row of is fl or0 -1 if each element of the ith row of (") is -1 or 0 Matrix (E) may be called the edge matrix.l)12 and n: the number ofjoint sets..] The next step is to compile a column matrix [.rn-n/zf (8. {E} : [e t €2 €3 €.:1 r(a.14)].l)l2l rows.15). The Edge Matrix In Chapter 6 we denoted the JP code.) :1 o if o. each of which corresponds to a particular i.1 l) :: 0001(a. The test conducted by these operations guarantees that any real edge of a JP is simultaneously in each of the half-spaces defined by the JP code. n(n .Computation of the Maximum Removable Area Using Vector Analysis In general. . we now form a testing matrix (D. in which if a. i pair.8] with cll-.14) in which 7 indicates the "matrix transpose" and 0 if in the ith row of (7) there are *1 and -1. :2 (8. i r1-6r.'.(1)(D).

I.2 . (: fi. 0100 has edges _lrr. r0 0 +l t. os or an) : 1031. An element value of *1 means that the respective JP has an edge directed by the respective Iry.(D) : [. :0101. ' 0101 has no edges.Irn. the JP has an edge pointed in direction -lui and if the element is 0.r. -1 0 -l -1 -t e2 0 as ao) ir. 0000 has edges -Irr. a2 ct3 an) : 0000. . . and . -I. ll 000\ I o 0l (D): f . the matrix (/) was giyen in Table For (a.(r) : -l I LI 00 0l r0 il and ". * fir) and each column to a particular JP. for (a. -Irr.11).4. According to (8. 00 /.(D) : 01 I and {E}: Similarly. and lrr. For the JP 0000. and -frn.i -l 0l 0 (/). g Gomputing the Edges Consider D" -. o r 0f \o 0011 For the joint system of Table 8. Block Theory for Tunnels and Shafts Chap.r. l. Accordingly. If the element is -1.r. Each row of these tables corresponds to a particular edge Ir. 0 I 0 -1 -l 0 (/).(a. . *l o 0 0 The edge matrix {E} has been computed similarly for every JP code (at az q3 ao) and entered as a column in Table 8. 1011 has edges Irr. the Jp has no edge parallel to I.3. _lrr. 0 I (/).? 0 ll\t -r il [E]: and I .

that is.4 does not show the edges for JPs with a repeated joint set.4 shows no edges for 0101 and 1010.f. it is necessary to flnd the projections f! and fl of its extreme edges t.. the edge matrix for 1031 was shown to be (0 1 0 I 0 0)r. (A JP with one repeated joint set consists of an angle in some plane through the origin. These then corre- spond to tapered blocks.. 1'.4(b) Edges of JPs That Are Cousins to Those in Table 8.4(a) Edges of JPs 0001 0010 0011 0100 0l0l 0110 01ll Itz -l 00 I -l 00 I Ir: 010 I -1 0-l 0 Irc ll0 00 0 -l -1 lzs -10-l 0 -1 0-l 0 lz+ 0 0 -1 -1 0 o-1 -l Is+ -l -l -l -l 0 00 0 TABLE S. 8. e2 a.5. and I23. This result means that the cousin centrosymmetric block pyramid with opposite edges. This is apparent. For the joint system of Table 8.JP.Thelimiting edges will be established by a column matrix {.4(a) 11il ll10 1101 1100 101l 1010 1001 1000 Irz 1 0 0 -1 l0 -l Itr 0-10-l 10 0 It+ -1-100 00 I lzs 1010 10 0 lz+ 0011 00 1 Ir+ 1l1l 00 0 Table 8. Therefore. Compute the value of IYt defined by I'lt : sign [(lu x A) ' lkrf (8. the values of IIll are given in Table 8. and t.. the extreme edges are projected along d. the only edges of this JP are f . Limiting Edges in the Tunnel Section To determine the angular interval and maximum removable area for a.") Note that Table 8.16) where lit:ftrxfi.11). lf a.a(a) and (b). ur" orthographic projections of I.2. by comparisonofofaidentical block pyramid is a Table 8. . its cousin has edge matrix columns in -{E}. has edge matrix [E]. For example. but the method is the same. for the example. and. and therefore it has only two "edges.Computation of the Maximum Removable Area Using Vector Analysis TABLE 8. as seen in the tunnel section (Fig. 1.8} determined by the following operations. and t. in ttrr plane perpendicular to d.

S) is an antisymmetric m x m matrix.0 . az a. {E} : le. 5.5 Values of t{f : Sign t(iii x a) . (8. The matrix of + t. where m:c?-n(n-l) 3. For every JP code (a.). g TABLE 8. tril lu h Irz frr rr+ Izs Iz+ lt+ Irz -1 rrr 0 -1 -t11 -l It+ -l 1 0 ll Izs -1 0-1 -l Iz+ -l -1 10 fgr -l -l -t -1 -l 2. I 8) 00 . . From [E] we define a diagonal matrix (. we provide the mechanism to assure that the normals to limiting edges are computed using real edges of a given Jp. (Note: In this step. an m x m square matrix . determine an edge matrix [E] as shown previously. and 0 of Table (s). (8.) Define a column matrix {^B} according to {B} : [bt bz bs b^f.17) 4. that is.5 is denoted as -1.258 Block Theory for Tunnels and Shafts Chap. €2 e*1. _l ll \_i _l _l _l _l 'l Matrix (. Compute the product (/) (s) (/).I is also an m x m matnx.l (s):l I _i _l .0 '0 (8. 0 -1 -l -l -l l\ lr o-r I t.1e) .I) according to 00 .gm . g.

If so. at ez e3 ct4 :0111. Then.r X ri is the normal of one of the limiting planes and I.11 to 8.t4.. least one element L -1 if the rth row of (/)(S)(/) includes only -1 or 0 and at least one element ./ Example 2. I. o1 clz o1 o4: 1031 .. using 8. [E]: (000 -1 .4.1)'.[E]: (0 1 0 0 0)' 0 00 0 0l (rxsx/) : 0 00 00 [B]: -l 0 0 00 00 lil [j) . x d. /00 00 fo o 00 lo o 00 (rxsxr):lo 0 00 [00 0 *1 I ill (B): (j) \s0 01 ?-. is one of the limiting edges of the JP.Computation of the Maximum Removable Area Using Vector Analysis if the rth row of (/XSX4 are all 0 or if the rth row of (/)(S)(/) includes both 1 and -L : if the ith row of (/XSX4 includes only 1 or 0 and at where b. {E'} . This normal was determined by Irr X d. considered as a candidate to be a limiting edge. from Table 8.. L Note: In the operations above we have established a normal to an edge. I1. from Table 8.(1 0 -1 -1 0)" 00 1010r 00 0000\ (/xsxr) : -1 0 o o t ol {B} : 00 o o o ol' -1 0 -l 0 0 0f 00 0006/ ril Example 3. we determined if all the other edges belong to the same half-space of f. Q1 a2 Q3 ct4:0010. Example fl:4. by taking the dot product with each of the other edges.4.

1li1 It+.7. Irr X d determines a vector normal to the projection in the tunnel section of a possible edge of JP.lzs 0111.Is+ 0100.. Consider a JP code having a limiting edge f.) (8. These results are summarizedin Table 8.20) is the normal vector of a plane passing through tunnel axis 6 and the limiting edge I. so 0011 and 1100 generate infinite blocks in the walls and have no limiting edges in the section. 1001 It+. From the previous analysis we learned that 0101 and 1010 are tapered. 8 Computing the matrix [B] for each JP code produces Table 8.4.13).. it is 1 when - -fry is a real edge. is a limiting edge of the projection of JP to the tunnel section plane: Ii. Therefore. The tunnel axis is in 0011 and 1100. each column of which is [^B] for both cousins heading the column.Irr 0u0.lzs 0001. tbis edge is repre- sented by er (the kth term of {E}) for the JP and therefore by bo(the kth term of tB]). TABLE 8.t x d) is a normal to the projection of a real edge. because these JPs have no edges.1101 lzt. The term e* is f 1 when I.7 Limiting Edges 0000.260 Block Theory for Tunnels and Shafts Chap. and it is 0 otherwise. The term bo is f 1 when all the other edges are in the . in the kth row of Table 8.10ll Irz.11 and 8. Iz+ Gomputing the Angular Intervals If I!. TABLE 8.r. 1110 Irr.i X d : Ilu X 6 : tnfu) or tn@.6 Limiting Edges of JPs 0000 0001 0010 0011 0100 0101 0110 0lI1 At AZ A3 A4 1111 1110 1101 lI00 1011 1010 l00r 1000 ln 0000-1 001 Irr 0 -1 0 0 I 0 0 0 Irt 0000010 -l r00000-1 Izs 0 lz+ 0010000-l Is+ 0t*I 00000 7. It defines a limtt plane of the "IP (see Figs. The limiting edges of a JP projected to the tunnel section correspond to the nonzero terms of [BJ. 8.6. is a real edge. In Table 8.1000 Itz.6 the limiting edges of each JP having removable blocks are the rows with *1 or -1.Is+ 0010. er(l.

1 0110 llaXd -Izs X 6 0111 -lrz X d -Iz+ X d 49. x ri) is nonzero for two vectors. These results are listed in Table 8.1 1011 Irz X d. -ltt X d 2s0.' oo " TABLE 8..2 24.6 0001 l13Xd It+Xd 204.1) that the tunnel coordinate system is determined by coordinate vectors fto:2 x d.2 olrt: fu x fit. .7 49.t x il. for each JP we haye now determined a vector in the tunnel section that is normal to each limiting plane and that points away from the JP.2 204.r 0100 -I12 X d l13Xd 70.8.6 1ll1 -lr+Xd -l23Xd 351.1 250.r 70. Then -boeo(|. fF^Dr.7 229./o: d x (2 x 6) and 2o: d. 229. The vectors i i'" " i .1 265. From Tables 8. The term 6o is -1 when all the other edges are in the half-space -e*(l.6 171...r 1r01 -l2a X d Is+Xd 265. Recall from (8.6 351. each normal to a limiting plane and pointing out of the JP.13 Lower-focal-point stereographic projection of lines shown in Fig.Computation of the Maximum Removable Area Using Vector Analysis Tunnel section Figure 8.7 0010 lz+xd X -l3a d 85. same half-space of f eoQ.8 Angular Interval (deg) for the JP Outward Normals to the Limit Planes in the Tunnel Sectiono From: To: 0000 IlaXd IzsXd t71.3 23s.8.3 5s.4 and 8.2 1000 InXd lz+ X d.3 1001 -\+xd lzs X d. r x A).7 r 110 -lrt X d -hrxd 24.1 85.11. 235.

8 Dr : Irr x d of Table 8. and 0 : q. is less than 180".21(b) Key block.1(b) Removable stable block under gravity 0101 None None Tapered block.14(a) 8.21(a) 8. The Maximum Removable Area and Maximum Blocks Having determined the angular intervals for each removable JP code. for each Jp.16(a) 8. for each vector of Table toward! 8. tunnel axis belongs to JP 0100 8.20(a) 8.) makes an angle 0.17(a) 8. y. measured from. the maximum removable areas are now fixed. such that 4r . order the angles 0: 4. These are shown in a series of figures keyed by Table 8.19(a) 8. then.22(a) 8. JP is empty 0110 8.15(a) 8. Compute lr:tr.8 therefore lie in the ftogo plane and Xt:Ilr.20(b) Key block 1110 8.18(b) Key block 1010 None None Tapered block. The angular intervals computed for every JP code using this procedure are listed in Table 8.1e(b) Key block 1100 None None Infinite block. JP is empty 1011 8.262 Block Theory for Tunnels and Shafts Chap.15(b) Removable stable block under gravity 0011 None None Infinite block.Xo (8.17(b) Key block.9 Figure Showing Figure Maximum Showing JP Removable the Maximum Summary of Removability Code Area Block and Importance 0000 8.14(b) Removable stable block under gravity 0001 None None Absent because of the tunnel shape 0010 8. from yo (measured from !o toward . The maximum finite blocks have also been drawn to show TABLE 8.4r. almost stable under gravity 0111 None None Absent because of the tunnel shape 1000 None None Absent because of the tunnel shape 1001 8.'90 Each vector rlr : (xr.9. volume is very small 111 I 8.22(b) Key block . tunnel axis belong to the JP 1101 8.

(a) Figure 8. The most important blocks ao= ? (a) (b) Figure 8.. Visualizing the maximum blocks in relation to the tunnel permits conclu- sions to be made about which blocks are key. .15 (a) Maximum removable area and (b) key block for JP 0010. Table 8. d rises from the paper (i.e.9 summarizes the remov- ability and relative importance of each of the JPs. that is.14 (a) Maximum removable area and (b) key block for JP 0000. looking south in this case).Computation of the Maximum Removable Area Using Vector Analysis 263 their intersection with the tunnel cylinder. the view is in the direction of -6. In these figures.

' (b) Figure 8.16 (a) Maximum removable area and (b) key block for JP 01fi).. . Figure 8.15 (Continued) \.

Computation of the Maximum Removable Area Using Vector Analysis 265 (b) Figure 8. . (a) Figure 8.17 (a) Maximum removable area and (b) key block for JP 0110.18 (a) Maximum removable area and (b) key block for JP 1001.

Figure8. .18 (Continued) (bt Figure 8.19 (a) Maximum removable area and (b) key block for JP 1011.

Figure 8. .Computation of the Maximum Removable Area Using Vector Analysis 267 (b) Figure 8.20 (a) Maximum removable area and (b) key block for JP 1101.21 (a) Maximum removable area and (b) key block for JP lll0.

2l (Continued) Figure 8. .22 (a) Maximum removable area and (b) key block for JP 1111. Figure8.

1301. The normals of the limiting planes that are pointed away from the JP are for l03l: -l* x d and lrt x d for 3l I l: -l* x d and -Itn x d The angular intervals are for 1031 : 235. each having one repeated joint set.6" to 24. 1101. The Angular lnterval for a JP with One Repeated Set Previously. 10L1. 1L31. 8. as shown above. The angular interval for 1031 is larger than that for 1001 or 1011 because JP 1031 contains the latter. [E]:(1 0 -t 000)' and [B]:(l 0 -l 000)" With one repeated joint set. for JP 3111. This suggests that blocks 1031. and 1131. (e) Figure 8.Gomputation of the Maximum Removable Area Using Vector Analysis (with no repeated joint sets) derive from JPs 1001.2iil (a) Maximum removable area and (b) key block for JP 1031. and 1111.23(b) and 8.3' to 70.24(a) show the maximum area for these JPs and Figs. .24(b) show the maximum blocks. JP is an angle in the plane of the repeated set and a nonempty JP has only two edges. In general.E'} has only two nonzero terms. w€ computed the edge matrix {E} and limiting plane matrix [B] of JP 1031 as {E}:(0 l 0 I 0 0)' {B}:(0 r 0 -1 0 0)" Using the same method. are also potential key blocks.2" Figures 8. {.1" for 3111: 351. Therefore.23(a) and 8.

Figure E.23 (Continued) Figure t. .24 (a) Maximum removable area and (b) key block for JP 3111.

or from the coordinates of the axis vector d. 8. the radius.2. of a JP. d is the angle AIA' and AA' is the length of . To demonstrate the methodology.26.25 Lower-focal-point stereographic projection ofjoint data in Table 8. 2. we can draw the limit planes of the maximum removable area by the following procedure. Figure 8. as shown in Fig.L From Fig. and .8.2.Computation of the Maximum Key Block Using Stereographic Projection Methods a block with a single repeated set has alarger angular interval than one lacking a repeated set. r. Plot the projection A and A' of the tunnel axis vector d. For each edge. assume that d is upward. using the system of joint sets presented in Table 8.25 shows this projection.26. Constructing the Limit Planes of a Given JP 1. Given a JP. of this circle is AA' ' zsind (8. draw a great circle through A. 11'.22) where. first plot the joint circles and establish the JP regions.2) to (3.t1 Figure 8. A'. using equations (3. Of course. f.7). This can be done from the given bearing and plunge of the axis. and -d. COMPUTATION OF THE MAXIMUM KEY BTOCK USING STEREOGRAPHIC PROJECTION METHODS All of the results of the preceding section can be achieved using the stereographic projection. a block with a repeated joint set has a greater resistance to sliding than one lacking a repeated joint set.

and an edge of the JP such that only the edge touches the circle. 8 -5li. the two dashed circles touch extreme edges of JP 1101. Figure 8. 8.8.27 shows this construction applied to corner 1 of JP 0000 for a tunnel in the direction of vector (0.5). and .--\ ---L- AA' 1 2 s'rn D- Figure 8. In Fig.866.27. 8.28. A'. in Fig.26 Great circles through A. JP 1031 (an arc of circle 3) is touched at its two end . A'. In Fig. The same procedure can be used for JPs corresponding to blocks with one repeated joint set. 3. These two circles are envelope planes of the JP.29. the two dashed great circles were drawn through A and A' for extreme corners of 0000. Block Theory for Tunnels and Shafts Chap. Figure 8.I. For example. the segment from A to A'. For each JP we can construct two circles through A. 0.27 Limiting great circles of JP 0000. 0.

29 Limiting great circles of JP 1031.Computation of the Maximum Key Block Using Stereographic Projection Methods / I I I 'il I I \ \ \ ------ -'z// ft Figure 8. / I I K-__z \ I I .l/ \I/ \-._--.fA. .-.28 Limiting gteat circles of JP 1101. // Figure 8.

using the stereographic projection entirely.30 Limiting great circles of JP 311l. using methods discussed in Chapter 3. Alternatively. Both methods are illustrated in the following sections. 8. . the extreme. 8. are the two extreme edges of the JP with respect to tunnel axis d). Similarly. derived in Section 4 of the appendix to this chapter. 8 Figure 8. For each limit plane of the given JP (each dashed circle in Figs. the two envelope great circles of JP 3111 are shown in Fig. the angular intervals can be calculated. 4. envelope great circles are the limit planes of the JP. .4r ) L80". Galculating the Angular Intervals from the Dip and Dip Direction of the Limit Planes The angular interval for the maximum key block corresponding to a given JP runs from 4r to r7. From the dip and dip direction and the orientation of d.30..21 to 8.30). measure the orientations of the traces of the limit planes in the tunnel section. The two v€lues of q are calcu- lated from the following equation. with the understanding that 4t-Ir<180" (8.23) lf tt. Block Theory for Tunnels and Shafts Chap. In each of these examples. points by the two dashed great circles. reyerse the subscripts. measure the dip angle and dip direction. corresponding to the planes normal to vectors (Iri x d) discussed in the preceding section (where f.

Determining the Maximum Removable Area by Stereographic Projection Having found the limit planes enveloping a JP. When the tunnel cylinder axis 6 is horizontal.25) The derivation of (8.24(a)1. planes b and c.B and C have been laid off in a view of the tunnel section looking south [with the tunnel axis rising from the paper toward the observer.B and C.32. respectively. 8. the maximum removable area can be drawn [compare with Fig. which lies between limit planes b and c in Fig. a1 and f r were computed from limiting circles. a. Similarly.f . so the results do incorporate the plunge of the tunnel.14(a) to 8. JP 1101 is outside the circle of limit plane c so that the block lies below the trace of this plane in the tunnel section. are seen to intersect the tunnel section at points -B and C. 8. In Fig.10. the maximum removable area can be determined graphically by finding the traces of the limit planes in the tunnel With the known shape and size of the tunnel section. drawn looking south). the limit planes.24) for JPs 0000. respectively. Since JP 0000 is inside the circle of limit plane b. The latter is the plane perpendicular to the tunnel axis.31. As a second example. 56" above horizontal from the east and 9o above horizontal from the west in the plane of the tunnel section. since 0000 lies inside circle c. the block lies above the trace of plane c. reconsider JP 1101. f r : dip and dip direction of the limit plane f : direction of d (Note that the angular interval formula is independent of the plunge. wnere ^ : I I if JP is inside the limit circle " 1-1 if JP is outside the limit circle dr. The JP lies above limit plane D since it is inside the circle of this plane and therefore it is seen above the trace of plane b in the tunnel section (which is. whose orientation does depend on both u and f. 8. The traces represented by . With this information. These limit planes project to the tunnel section at points . the maximum removable area is drawn [compare with Fig.) (8. 8. 1101. However. 94' above horizontal from the west. sin (B .) The angular intervals were calculated using (8. 8.20(a).25) is also presented in the appendix to this chapter. Similarly. and 3111. 4 : 90(1 + s) * ur sin (p . the block lies above the trace of plane b. . as in Figs. of the tunnel section.Computation of the Maximum Key Block Using Stereographic Projection Methods 275 4:90(l + s) * sin-1 [sin a. The results are summarized in Table 8. again.F)l (8. for JP 0000.24) may be replaced by a simpler expression. and the shape and size of the tunnel. d. (8. 1031.24) . and 49" above hori- zontal from the east in the tunnel section.

lr 8.2 4t 31.7 4t 8.1 Dip Direction of Limit Plane (dee) 163.12.3 267.8 25r. 8.0 I Limits el and 41 of Angular Intervalb (deg) t71.30 31l l 73.28 ll0l 55.5 246. see Fig.8 22t.2 I 235.2 I 235.-1 if JP is outside the limit circle.3 rt r or Ir 4t l 60. bThe angular interval runs from 4r to ry clockwise .O 2s9.27 JP Code 0000 TABLE 8.1 4t 60. t .6 -1 49.3 4t aWhere S: +1 if JP is inside the limit circle.9 -l 24.7 4r 8.1 4r 86.0 -l 35r.8 251.1i6nm Removable Areas Dip of Limit Plane (deg) 31.9 I 26s.7 .8 -1 70.f0 1r[g.1 163.29 1031 37. { i' o) Figure 8.

. Plane c \ -\--- --'\' Plane b (b) Figure 8.Computation of the Maximum Key Block Using Stereographic Projection Methods 277 \ \ \ Limit \ -\.31 lntersection of limit planes with the plane normal to the tunnel axis (looking south) for JP 0000.

Block Theory for Tunnels and Shafts Chap. 8 (a) (b) Figure E.32 Intersection of limit planes with the plane normal to the tunnel axis for JP 1101. .

California. Moreover. the discontinuities to be weaker and more numerous. portal construction is often burdensome and expensive. (b) in metamorphic roct.33 Photographs of portals of tunnels: (a) in diatomite. Figure 8. and the action of surface and groundwater more troublesome. Because the intersection of the tunnel surfaces and the ground surface creates complex (a) (b) Figure 8. near Lompoc. For these reasons.33 shows three examples of tunnel portals. . near the free surface the rock tends to be weathered.Removable Blocks of the Portals of Tunnels REMOVABLE BLOCKS' OF THE PORTALS OF TUNNETS Portals of tunnels present increased opportunities for block movement since each of the tunnel interior surface planes intersects the free surface.

" for theaccess tunnel to Kerckhoff II underground power house. . after initial blast following shooting of "relief hole. (courtesy of Pacific Gas and Electric Co.33(Continued) (c) In granite.34 SP and EP for the floor and face of the tunnel portal. Block Theory for Tunnels and Shafts Chap. 8 (c) Figure 8.) sP (wl) Figure E.

(see Fig. sin 223" . cos 60") rfl3:(0. Free planes W. 8. 8. the general problem is difficult. Corners Ctz+. -cos 158". For purposes of expla- nation and demonstration. Edges Err. cos 223" .34. -l) fir: (sin 158". 8.0.11 Planes of the Tunnel Portal Dip. cos 158".Removable Blocks of the Portals of Tunnels blocks with curved edges. the rock mass is in the lower side of this plane. and the removable blocks of each element. 0) The removable blocks of the portal of the tunnel are the blocks of the tunnel that have portal face planes Wt or W. Err. that contains r0r. as additional boundary planes. these have been labeled Wt through Wu. l) fr z : (sin 60' . roof. TABLE 8. 8. The vectors fi. SP. be the positive normal to plane W.36) We will compute the EP.34) . The Free Planes of the Portal The free planes of the portal are the interior walls.3a) .E .Thenby U(fi. In Fig. and W2 (see Fig. Crrn.35) . The planes are described in Table 8.a Dip Direction. sin 60' . Crrr. 0) frs : (-sin 158o. p Name (deg) (deg) Description Wt 0 0 Floor of the tunnel and the portal Wz 60 223 Face of the portal Ws 0 0 Roof of the tunnel W+ 90 158 Northwest wall of the tunnel Ws 90 158 Southeast wall of the tunnel It will prove convenient to adopt a convention that the positive direction of the normal to each face is that which points into the free spqce. Let fi. and floor of the tunnel and the free surfaces of the natural ground at the location of the portal. 8. (see Fig. are as follows: r?1 :(0. for a tunnel of rectangular shape with horizontal roof and floor. we will assume a polygonal tunnel section and planar free surfaces.) we mean the half-space of W. Free plane W'1is the portal base. and Ctr.11. Ern. then EP : L(fr r) and SP : U(frt) (see Fig..0. and. The elements of tunnel portals are: .

assuming a rectangular tunnel shape. and Wl as faces. Lfld Err-as shown in Fig. then EP : L(fr r) and SP : U(fir) (see Fig.35.282 Block Theory for Tunnels and Shafts Chap. respectively. Free plane Wz is the face of the portal. 8.tal There are four edges in the portal-.1are those blocks that have free planes W. Each complex block of edge 812 can be divided into two types of convex blocks with excavation pyramids. EPI : L(fr t) EP2 : L(fr r) . The removable blocks of portal edge E.Erzt Eztt Ey.35 SP and EP for edges of the tunnel portal.34) The Edges of the Por. the rock mass is in the lower side of this plane. I Figure 8. 8.

Then.28) SP:U(frr)uU(lfi4) (8.L(t1t4) (8. EP: L(frr)^rL(ti)3) (8.L(vDs) (8.U(fi2) (8. Err.36 SP and EP for the corners of the tunnel portal. u EP2: L(frt)u L(1fi2) (8.Removable Blocks of the Portals of Tunnels 283 Figure 8. so for edge E2n.32) and SP : U(fi r) u U(tt) s) (8.3) (8. the criteria for removability are JP+ @ and JPnEP:@ or JPcSP where EP: EP. and E2s are convex edges.2e) For Err.31) For edge -82r.27) Ern.26) then sP: -EP:U(fi). using Shi's theorem for nonconvex blocks. EP: L(frr). EP: L(frr).33) .30) and SP: U(frr)uU(ti.

and 1101. EPr : L(frr) n l. Wy and Wo as faces.(fi3) (8. The portal is a critical position of a tunnel.and C2rr. and edge E2. the criteria for removability are applicable if EP:EP' U EPz:L(fr") n(Z(#3)UI(la4)) (8. 8. respectively. in the portal as shown in Fig. 8.) n U(tn/) (8. the removable blocks are 0000. Figure 8.35) Then using Shi's theorem for nonconvex blocks.36. From Fig. Cr"n.284 Block Theory for Tunnels and Shafts Chap. Crrr. Crrn. 0010. is the most critical element of the portal. 0001. 1011.37 we can see that the EP of edge E* is very small. 1001.35).34) EP2:L(frr)nL(fi4) (8. the SP is very large and there are many removable blocks. 0011.36) and sP : _EP : U(fi) v (U(ti. with excavation pyramids. 8 1111 Figure 8.37 shows the EP and the removable blocks of edge E2r. Figure 8. The Gorners of the Portal There are four corners.Ezr (see Fie.. Each complex block of corner C23a c?flbe divided into two kinds of convex blocks.37 EP and removable blocks of edge . 8. Accord- ingly.35 shows the EP and SP of these edges.37) . The removable blocks of portal corners Cqr arethose blocks having free planes W.

Each complex block of corner Cr"n can be divided into two kinds of convex blocks with excavation pyramids.Removable Blocks of the Portals of Tunnels 285 Similarly. are different from corners C2rn and C2tt (Fig.45) Figure 8.38 EP and removable blocks of corner Czr+ (see Fig' 8.44) SP: U(fr') n(U(lfir) u U(las)) (8.r.3) n U(f/s)) (8.36 shows the EP and SP of these corners.0010.36).fu""' .3e) Corners Cr2n and C12. 8.(fi4)) (8. Figure 8. respectively: EPr : f(fir) (8.) n f. . for cornet Czss we have EP : L(frr) n (L(fi ) u L(lAr)) ( I I 'ffi-. the criteria for removability are applicable if EP : EPl u EP2 : L(frr) u (L(fi.42) sP : -EP : u(fi) n (u(tir) u u(fi4)) (8. for cornet Crzs. Figure 8.43) Similarly. EP: L(fr') u (L(1fi) n L(fis)) (8.36). 1111 .38) SP: U(frr) u (U(f. 101 t t I I \ I I /.4) (8.38 shows the EP and removable blocks of corner Czsni the removable blocks are 0000.4r) Then using Shi's theorem for complex blocks.40) EP2 : L(fr r) n L(fi.

8.0010. 0001. 1001. 0001.12. we can compute all of the removable blocks of each position of the portal. 0001. 011 I Ezs u(fr) u u(fis) 0000.ffi. Summary of All Removable Blocks Using formulas for the EP and SP for each element of the portal. 0010. 0010.0001. 1101 . 0011.8 1111 . 1000. 1101 Ez+ U(fri u U(fi. 0010. 1100. 1001. 1101 Cn+ U(fr) n (U(r?z) u U(fr+)) 0010 Cns u(fr) n (u(fr) v U(r0s)) 0010 Czt+ U(fr2\ u (U(fir) n u(fi+)) 0000. 0001. 1001. as established previously.0010 Wz U(fr2) 0000. "'l 011 1 i li ti 0011 Figure 8. The removable blocks of each element are shown by Table 8. Figure 8. The only removable block belongs to JP 0010.286 Block Theory for Tunnels and Shafts Chap. 0011. 0010. 1011.36).0011 Czts u(fr) u (u(o3) n U(frs)) 0000. TABLE 8.4) 0000. 0010. and 0011.39 EP and removable blocks of corner C12a (see Fig.39 shows the EP and removable blocks of corner Crrn. 0001.0001 Erz u(fr) n u(fi2) 0010 Ezs U(fii v U(frs) 0000.12 Removable Blocks of the Portal Position Space Pyramid (SP) Removable Blocks lvr U(frr) 0100.

Because a+ IP: A U@. Suppose that JP : hu@. Suppose that there is a removable block having joint pyramid JP. Proof. Proof. is the normal of plane f.) (11) we can find an i such that a + u(fi) (12) a. From the proposition EP: U(fr(lJ) u U(ir(0^)) (3) and EPnIP:@ (4) But U(fr(O )) = *a (s) u(fi(|-)) = ta (6) where "=" means "contains.) (10) where fi. then JPnEP= 6or-d (8) therefore JPnEP+g (9) But this is a contradiction since this block cannot be removable. there is a removable block in the tunnel with joint pyramid JP.Chap. JP is a removable block of a tunnel if and only if +a + rP (1) First we prove that if de JP or -deJP (2) then JP is not a removable block. Next we prove that if +A + JP. 8 Appendix 287 APPENDIX PROOFS OF THEOREMS AND DERIVATIOIUS OF EOUATIONS 1. poioliog into the joint pyramid JP. PROOF OF TUlIINEL AXIS THEOREM Axis Theorem.fu <0 ( l3) ." So EP: U(ft(?r)) v U(n(0*)) = *a (7) ff +A or -d e JP.

f11 :0 Q9) "rz' rrr x n) nrl :-'::.(4.fir| (24) Suppose that mr: (f.fti (20) I.r : (h x fe).).).i (le) I*r : (fio x fi.sign[(. C : B x C .(fi.sign [(fi.U(nk) (18) has three edges vectors (see Fig. 8.A) .nk(frt'il] (28) mr. x fi).40).fir] (23) It* : (4 x fi*)..sign I(fi.U(ft). x fir).). Expandingr we have mr: ftt(fi.d)l(fi.l(firxfi.sign [(fi.).) x .) (14) we can find a 7 such that -a + u@) (15) -d-n1 { or O (16) Choose fi* such that k+i and k+j (17) The pyramid JPo: U(fr.6\1.sign [(f. x fi. Block Theory for Tunnels and Shafts Chap. B.sign l(fr..sign [(fir x fi. x fi)'fir) (22) l*t : (fir x fir).sign [(fit x fi). 8.)*l.:*r(n' From (13).firj f. we have mr.. A: C x A. then Itr : (fi. frr : lfi{fi.. x fi).fi1| . x fi.40): : (fi.sign [(] (2r) SinceA x B.(ftr x fit).fr1 (0 foranye>0 (30) Itrr .fi{fi. x fi).fir[ : (fi..ffiI :.Tiffi \?!.. x fr). x fir).fir| : (fit.illnt.A) e7) IIr2 : -e[fi{frk'il .il . 8 Because -a + JP: t:1hu@.0. X fi.6 (25) and x fi).)xdl (26) where e is a small positive number (see Fig. x fi).

Chap.sign l(ft. * ft.(fit x fip).ill.sign l(ff. 'fro ( 0 (32) From (29) and (30). x fi). x fit).sign l(fi. From (12).d) * nlfrt.fps(O (31) nr .+Dz). x ft)'firl : -(frt'il\@t x n)-frl..fio| : -(fii.I.fio1 From (16) we have m.il1(!. 8 Appendix 289 Figure 8. Irr : lfit(fi.4. we have m1 .r(0 (33) .fiel.).0 Projection of vectors introduced in proof of tunnel axis theorem. we have (m.

(0 (34) (mr*<e<q.. @. i : 2. respectively. PROOF OF PROPOSITIOIU ON ANGUTAR RELATIONSHIPS FOR AN EMPTY INTERSECTION u(n@)) n JP .) (43) Let the edges of the JP be lr.A(0) n JP .290 Block Theory for Tunnels and Shafts Chap. andIi. Accordingly.then there are values a > Oand D > 0 such that n(il: afi(q. from (31) and (32) we have..@ (36) Because JP c JPo (37) therefore U(m. we find that mr:(ft.1...+mz).Ii"(0 (3s) then U(m. (m. I!. Ir : 0 n(n). (a) If 4110 aq.] If ttr] 4. 8. 8 Choose 6 small enough.tlt < 180" (45) .)xdl then (mt*mz). * mz) n JPo .. then.. add 360" to tl.g if and only if 41 10 1rt.. fr<0. Ir : 0 fi(tD. and fi(4) are shown in Fig. fr(tt). (B..14..If .lr..*mz)nJP-O (38) Recalling (25) and (26)..11).Xfir)Xd (3e) 12: -el(fi*xfi. | [The spatial relationships of I..2. The rotation from Q@)to QQD is clockwise.1 (44) fi(n).. 2..d:0 (40) We can find 01 : 0^ such that fi(?r): fr(|-): nr * mz (41) (u(n(o))u u(fr(o-))) n JP: @ (42) From the proposition there is then a removable block in the tunnel with joint pyramid JP.z) (see Fig..3.The extreme lines of the orthographic projection of JP in the tunnel section are I'. Ir ( 0.2). . i: 1... 8.) * bn@.3. so that 0 t1.

.that points into the block. U(6.12) (54) ^ where d..3. Ae U(O.If a ( 0. rr (46) n(il. + bn(o)' rt : aff(tt). then 41 10 < q..-..)' I. i:|. . rr : afr(q') ' r. 8. fi(O)' fr: ah(n) 'f.Chap. i: 1. U(fr(0))nJP+@ contradiction (53) 3.. {u(-fi(tt). Proof of (a).. is the normal to plane P.: u@(0)) n JP . such that A a B.1. then . * bn@..I. then n@) : aft(ry) + bfi(tt) (50) where a:i0 or D ( 0.).| (47) n@)' \ : afi(4.) that passes through Qu and TB is the rock mass outside the tunnel. * bft(4)'It : bfi(ttr).\: afi(Tr)'r. using (44).B belongs to the maximum removable arca of JP.Q. Let B : R (u(6.) 'r.A)-U(6... f.).I. * bfi(tlt)'\ : ( 0 bfi(qr). Q)is thehalf- space U (6.Q. Q(il)} (8.B is a removable block of JP in a tunnel.B of JP such that the projection of B along d is exactly the projection of the maximum removable area of the JP. and d. QQt)) n U(-fi(4)... ... B c.Z 0 (51) Similarly.n (56) . n(o)' \ : afi(4.e)) TB (see Fig. using(44). ( 0 (48) Since n(0) . Suppose that 0 is not between 4t and 4t. It ) o (52) Since JP includes /.@ (4e) (b) If U(ft(0)) n JP . I bfi(tl)' rt : afi(q).8) (b) There is a removable block . and JP is contained between f t and I.)' r. PROOF OF THEOREM OtU THE MAXIMUM REMOVABLE AREA (a) If . ( 0 for all i. ( 0 for i: 2.fl (55) and U(6. For any point l. namely. if b < 0..a. I Appendix Applying (44) to (43) gives us n@).

and 9o. Xo:2xd yo: dx(AxA):Qfrxt)xd (8. 4..9o. 41 10 l rtt (61) Let Q(|r) and Q(?r) be limiting points on the boundary of B(A). The angles 0 : q.)..) n u(. 8.g. and fi. QQD)} along d is shown in Fig.12 as region B(A) with apex at A.t) (s7) Let B(A): . from !o (toward io) to fi1 or fi. This projection is also the projection of JP after moving its apex to A. Block Theory for Tunnels and Shafts Chap. compute ft'. DERIVATION OF EOUATIONS (8.) (60) so ^B(l{) is also a removable block of JP. o.20). and ft.0 U(0. 8 hu@. A)) .fi. The projection of {U(. B is exactly equal to the projection of the maximum removable area of JP. delimit the angular interval (see Fig.e1) n Er (62) ": also has the same projection and if A is at the intersection of I'. JP n U(fr(0)) .{ is any point in B. are in the *o!o plane. Q@)) A U(-fr(tt..8) Proof of (b). their normal vectors fr.1) zo:d' The limit planes pass through d : ?oi therefore. A) n TB (s8) then B(A) c B (5e) [Figure 8. In the boundary of the tunnel.. B.(A <0.25) Recall the tunnel coordinate system (ft'. that is.12 shows the relationships between B(A).. by virtue of (60). as Then if . We have previously found the dip and dip directions of the planes normal to fi.In order to compute the angular and I' The removable block (n u(fi. Then by the proposition on angular relationships for an empty intersection. A(l) is a pyramid with apex at A and having the same shape as the JP.. I'.. if there is a point Q(0) e B(A). 8. B(A) c {U(-n(q ). and 0 : q. and II.l Both B(A) and ^B have the same JP. Q@D} (8.24) and (8. JP: hu@. . Q(q.12).

8 Appendix From (8. fir'fto: S(sin a.cos(B ..ft). sign (.ft) *cosdcosdr:0 then cos (P . * cosdcosa.f t) (67) From (64) and (66). cos d) : (-sin d cos B. and fi. we get ':illl i.lo:Scosssinfircos(P-F)-Ssin&cosdl (68) Because d is in the limit plane and fit is the normal of this plane.)'.by fi.. Knowing that the normal plane of fi. sin B1.S sin a. fir. : g . o (6e) because 0 { a ( 90". fir.sin dr cos Bl sin B) or frr. -Ssindcosdl or fir. ''-io ) o (71) [180-sin-l ( : -..] :6 sincsina. * sindcospsindr cos F. sin p1 cosf . sin a) (64) 2o :4 : (sin a sin f. sin a sin p. sin d cos p.sin-l (fir.i. -S cos dr) (66) I wnere od: t 1 if JP is inside the limit circle 1-1 if JP is outside the limit circle From (63) and (66). cos a) x (-cos B.f')l .o: (sin c sin f.0) (63) yo : d x ft.'ft. From (68) and (69) we know that sign (fit'.i. has dip c. 1) x (sin a sin B. cos P.lo: Scos a sin p sin a.. sin (B . -cos d cos sin 8.0) so fto : (-cos /.F): -ff#ffi .9).ar--fo)) * sign (fir..S (70) Denote the position angle of fi. sin d cos B. < 90".lJJ#r. 0.then o : {ltl^-'(fi'. in plane xoloby q. * Scos d cos B sin u. sin p.0) so I o : (-cos a sin B. -S sin dl cos f .S[sin usin p sino. xo : @ x d): (0. then fir : (-. sinp. fto - -.ft) (72) Substituting (70) and (67) into (72). then d'fi'. and dip direction B.. sin f .S sin o.1). 0 { o. cos a) (65) In the following we denote both fr.:i .\ol-L tl :90(l . sin c cos B.

sin (p .::::* = sin(/-F):l or -1 Assume that d is upward. 4:90(l + s) * sin-'lsincl. . cos a) The formula for calculating the angles to the limit planes of JPs intersecting a horizontal tunnel is then 4:90(I +s)*drsin(f . then -d is downward.sin-l (sin ct) - t' * dr'sin (P .2s) where o: I L if JP is inside the limit circle 1-1 if JP is outside the limit circle . finally. When d : 90".F)l 90(l + . Block Theory for Tunnels and Shafts Chap.F) Q4) because o ( c. I The formula for the angular interval is.sin (P .fr) (8. Given the dip a and dip direction P of the plane normal to d.. then 6 : (sin a sin f.B.S) * sin (P .f'):0 so sin(p-F):l or -1 (73) Formula (8. (69) becomes cos(B . sin d cos p.24) A horizontal tunnel is a special case with d : 90o.).24) then becomes 4: 90(1 + s) * sin-'[sin d.f ')] (8.

as noted in Chapter 1. Certain of the blocks created by the intersec- tion ofjoints and excavation planes prove to be finite and nontapered. it is a subject worthy of further consideration for weak rocks or for underground excavations at great depth). This is also true of tapered blocks (type IV). which are in a position to slide or fall when the excavation is made. Stable blocks. chapter 9 Tho Kinornatics and Stability of Removalole BIoCKS The previous chapters produce methods to determine the removability of a block intersecting a free surface. but will not because the available friction on the faces in contact is sufficient for equilibrium III. Joint blocks behind the free face have no space into which to move until some surficial blocks vacate their positions. which cannot slide or fall even when the angle of friction on the faces is zero because the orientation of the resultant force promotes stability Only removable blocks merit stability analysis. This possibility is ruled out in this book (although. Infinite blocks (type V) intersecting the free surface can move only by becoming finite through rock fracture. In Chapter 4 the removable blocks were divided into three types: I. unless support is provided II. Key blocks. which can be expected to move when the excavation is made. Having identified the removable blocks as candidates for further analysis. it is convenient to effect their subdivision into the three types on the basis of . We have called these removable. Potential key blocks.

1. on an incipiently sliding block. LetA.f. The direction of the resultant force must be specified but no joint-strength prop- erties are needed.8. or real key blocks (types II and I).1) We assume that there is no tensile strength across the joint. Thus F can be used as a vehicle to discuss the limiting conditions. We shall denote a specific removable block by B. a negative value of F implies that the block B is safe from sliding. Goodman (1976).B cannot be expected to be exactly in a con- dition of limiting equilibrium. if any. Under a given set of forces. MODES OF SLIDIIUG In this section we establish relationships connecting the direction of the resultant force. . The Kinematics and Stability of Removable Blocks Chap. Forces Acting on B There are three contributions to the forces acting on block . as shown in Fig. and the direction of sliding. Limit equilibrium stabilityanalyses have been publishedfor tetrahedral blocks by Wittke (1965).Ignoring rotation. (e. directed into block . a mode analysis is performed to distinguish stable blocks (type III) from potential key blocks. in order to separate potential key blocks (II) from real key blocks (I). Coupled with other kinematic constraints and a specific direction for the resultant force.o. in which motion ensues without acceleration. The results of the mode analysis guide the stability analysis. The unit vector of the direction of sliding shall be termed . so &>0 . be a unit vector normal to joint plane /. Conversely. We shall extend this established technique to blocks with any number of faces. To bring B to a limiting state. The discussion will imagine a state of limiting equilibrium. Londe (1965).9. Shi (1981). Finally. sliding mode is applicable to each JP. First. The kinematic mode analysis discussed in the next section has not been presented previously. and others. a sliding equilibrium stability analysis is performed. every part of -B undergoes motion described by the same vector. 9 two kinds of analysis. the block tends to slide unless artificial support is added. these rules will permit us to establish which. The resultant (N) of the normal components of the reaction forces on the sliding planes. Then the normal reactions are N:EN. we add a ficticious force -FS. John (1968).1.8. with friction angles input for each joint surface in contact. When Fis positive.

icJP (e. .1 Explanation of ficticious force F and sliding direction vector ^i. seepage forces or external water pressures. : g. The condition of equilibrium for a potential or real key block ^B is t+FN. Z > 0.8. sliding will occur if f.Modes of Sliding 297 :l Figure 9. the sliding direction ^f of remov- able block B belongs to the JP of block . 2.*F (9.2) and the resultant of this and the ficticious force ts -Ts : -t tr/. and support forces from rock bolts or cables.>0 From the theorem of removability. that is. by design. Chapter 4.4) For a potential or a real keyblock.0. Since positive F implies sliding.Fs (9.-Li:0 (9. inertia forces. The force r will be termed the active resultant.6) . 3. The resultant T of the tangential frictional forces: Tr : .3) so T:ElI. including weight. tan 0$ . The resultant r of all other forces acting on block B.N' tan $'3 (e.tan6.5) with f>0 and I/.

^f:jl where Sr:9t#+& (e. Sliding on a Single Face Figure 9.r that S : i.f is parallel to only one plane of B. 8:f (e. .8) I .5) becomes t:TS For a key block or potential key block. The condition for lifting is that .2 shows a key block translating freely from its home.7) The joint pyramids have been defined as closed sets.f is the orthographic projection of r on plane f. plane i. opening from each face. Figure 9. T > 0. Since no joint plane remains in contact. : 0 and (9. Since no joint is in contact. 9 Figure 9.3 shows an example of a block sliding along one of its faces. N.2 Lifting. If S is not parallel to any plane of a JP. Therefore.f cannot be contained in any joint plane. and the sliding direction . . In this case.f be contained inside JP but not in its boundary.5) . then the necessary and sfficient condition for B to satisfy the equilibrium equation (9. meaning that a JP includes not only the space inside the pyramid but the boundary faces and edges as well. Proposition on Lifting (Free Translation). We shall call such a mode lifting. The Kinematics and Stability of Removable Blocks Chap.

Modes of Sliding 299 / /. all of the joint planes except plane i will open and i c JP ) P. along their line of intersection) since that is the only direction common to both planes. which is proved in the appendix to this chapter.- \\ \s Figure 9.e. 3 : .APt (e. ^feJPnP.5) and (2. Sliding on Two Faces Figures 9.. where P.4 and 9. the sfficient and necessary conditions that the removable block B satisfy the equilibrium equation (9. and At' t < 0. P. the sliding direction is an edge of the JP formed by the intersection of planes i and j. In this case.e) Furthermore. Proposition on Single-Face Sliding..fir : #ihsign ((fi. is the orthographic projection of s on plane Pt[given by (9. represents plane i.10) .5) is i : S.a\\ t.) . The sliding direction ^i is the direction along the line of intersection that makes an acute angle with the direction of the active resultant r.7). determined by (2.8)1. where S. If the sliding direction S lies within only one plane.3 Single-plane sliding.5 show examples of blocks sliding along two planes Pt and P1 simultaneously (i. r) (e. x fi. This is stated formally in the following proposition.. and frt is the upward normal vector to plane i.

9 Figure 9.5 Double-plane sliding.300 The Kinematics and Stability of Removable Blocks Chap.4 Double-plane sliding. Figure 9. .

(9. THE STIDING FORCE The equilibrium equations for free translation. x fir).4'xxry: "-1ft1 *.11) and the equilibrium equation is TS:r (e. (e.4) becomes F:T (e.The Silding Force Proposition on Double-Face Sliding. If the sliding direction is simultane- ously in two planes. r€sp€c- tively. P.sign((fi. Using these equations.0. a o and s: . and Pr.^f. as given by (9. the necessary and sfficient conditions that block B satisfy the equilibrium equation (9. * F (e.xtl (e.5) are ir.14) It is shown in Section 1 of the appendix to this chapter [equation (2)] that the equilibrium equation (9.4) becomes T: Nttan6..16) Also. Nt :0 for all joint planes and equation (9. The proof is given in the appendix to this chapter.13) When gravity is the only contributor to the active resultant. equation (3) of the same section leads to Nt:-l'At (e. x r) x I. and P1.8).11) gives us r: lrl (9. r) where 3. Single-Face Sliding Since there is only one plane in contact.5) leads to TS: (0..r7) . we can compute corresponding equations for the sliding force FJ. single-face sliding.<0 a.15) so T:li. and 31 are the orthographic projections of r on planes P. and double- face sliding have been given in the preceding section.r2) and combining with (9. Lifting In this case. F is simply the weight of the block.

O)l : ((W' sin2 &.0. we have F:l0. (e. x rl : leW sin c.23) .W sin o. (e..Nr tan4..19) into (9.4) becomes T: NttanS. are a. * W2 sinz d. : (sin a. . \fr. x rl . 1fi. tanQt + F. x dr). sin Ft.. .20) (9. W cos d. .Nrtanfl (e... fi. and p. cos a.itanQ. removable block and all the other joint planes open.-W). Nr>O so F:T. fr'r(0 Substituting (9.14). rltan f. sin dr cos B.2L)expresses the net sliding force when the active resultant r is due to gravity alone.20) is the formula for the net sliding force of single-face sliding along plane Pi. cos 8. + il.22) From equation (5) in Section 1 of the appendix to this chapter. sin2 f )r/2 .@.le) Since from the proposition for single-face sliding.18) gives F:\fi.18) where At : -frt sign (fir .@' x 0t): -(r x 0)'(0' x 0r) glvlng m N. N..l : | -W cos q+l : W cos u.21) (g. W sinc. r) (e.(0.16) and (9.L7) into (9. \ft. The normal reaction force Nr :0 for all I + i or7 and (9..20). From (9. W>O Suppose that the dip and dip direction of plane P. When the active resultant r is given only by gravity. The Kinematics and Stability of Removable Blocks Chap. Double-Face Sliding In this B slides along planes of joint sets f and 7 case. Nd>0. r:(0.).: -(r x 0). (e. tan$. T>0. x 0i) (e. we have F: W sin a. r.xrl *0. cos2 f . 9 Substituting (9. sin fr.

866. and friction angles of four sets of joints.20).24) This follows from the proposition yr . (fi. finally.(fi. (9. then from (9) in Section 1 of the appendix to this chapter.22) gives. i z : (0. (9.l(r x ft)-(fi. x fit)ltanf. x firl Substituting (9.A)'@' x fitlz I ft)l (e.21) gives the net sliding force for a key block or potential key block sliding on plane P.24). Nt: lG x. 0.. F:Wfuilr. rr:Wir.23) may be written A. x fit)lln.27) Equation (9.500) TABLE 9.26) in (9. T: r. (9. . 1. ir < 0. .13).x ntl. dip directions.27). and P1. Given active forces. An Example of Sliding-Force Calculations Table 9. 0. and (9. we have calculated the net sliding forces along all of the sliding directions. The results are presented in Table 9. 7 > 0. p Friction Angle Joint Set (deg) (deg) (dee) | 7r 163 15 2 68 243 20 3 45 280 40 4 13 343 30 Using formulas (9.-1) and t2 : Wi r. so . Dip Direction.1 a Dip. x n)l rrr-W (e.1 gives the dip.xfi)l (e.(fu * fir) ltan Srl e.r _ l(r x fit).2s) 1fi.The Sliding Force 303 Since & > 0. as shown in the appendix to this For a key block or potential key block.0.2. Similarly.(fi.l(r x fr).25). and (9.. fr:(0.26) 1fi..

ool4/ t. as we shall see.\ . .6rw -0. . the block slides if it is removable.36W 213 -o. we can compute all of the sliding directions. '. If F is positive. the sliding directions sry are given by (e. 0. S : f .86W . Given r When the direction of the active resultant r has been determined.45W o.4 *0.36W -o.3 gives the results of a sample calculation.1 with r : (0.72W 214 *o.nw o. For sliding on a single joint faces (f).62W l.9 TABLE 9.45W -t.92W *- 3 O.866.e). For sliding on two faces (i and j). the block is safe. A "mode analysis" determines a complete list of JPs corresponding to all the sliding directions. are determined by (9.2 we can see that different resultant directions generate different net sliding forces. KIITEMATIC CONDITIONS FOR TIFTIIUG AIUD STIDING In this section we prove that there is only one joint pyramid corresponding to a given sliding direction and then establish procedures to identify it. For lifting modes.50w 4 *o.72W o. in the next section a given excavation configuration does not yield removable blocks corresponding to every mode.34W o. 3 -o. 0.62W 2 o:l9w 0. the sliding directions . This will be demonstrated in subsequent sections -----/ The Gomplete List of Sliding Directions. meaning that it may be a key block (type I). However.8). Table 9.l9w From Table 9.o3w 3r4 -o.35W lr2 o.l6w 1. corresponding to the joint system of Table 9. . The Kinematics and Stability of Removable Blocks Chap.f belonging to a given system ofjoint planes.oow I 0.IIW o. 0. meaning it may be a potential key block (type II).500).2 Net Sliding Forces (F) Net Sliding Forces (F) Sliding PIanes forrr : VI/ir fOr rz : Wfz None l.f. If F is negative.

.500) Vector ftr 0.8262 -0.866.9743 i 0 0.2113 3tz 0.34) provide sufficient information for identifying all of the corresponding d. . .8949 3z 0.2w7 0.29) to (9.2174 3s+ o.2r86 0. and.0657 0. 4. (9. i is parallel to P. l+iorj (e.7071 fit -0.6811 o. > 0 (e. Using (9.0599 0.325s ftz -4...Kinematic Conditions for Lifting and Sliding 305 TABLE 9.8041 0.9563 o.9745 -0. For the case of sliding in plane P.4233 0.31. are parallel.a. Then with (9.28) and the proposition on double-face sliding.34) can be replaced by <.. every removable block has a sliding direction along which it can be moved without colliding with the adjacent rock mass.1227 0.9042 0.i and r.0500 According to the definition of 'oremovability" given in Fig.6r< 0 ( 0.8261 fts *0.8977 jts 0. and (9.f+ 0.2185 -0.i. and .61{-0 (e.34) If none of the vectors r.33).0.5973 Stc 0.32) st. ^f is not parallel to any plane P. is the normal directed into the block.1487 .31.5587 3zt -0.6963 o. for each plane / of the block i.t6(a).9116 0.s000 .3890 0.9755 -0.1968 .4209 o.33) ij.1746 -0. For sliding on planes i and7.l+i (e.28) 0.2151 o.5196 o.6415 -0.2764 -0.28) and the proposition on single-face sliding. Then since d. It follows then that for such a block i e JP (9.5256 -0. (9.5755 -0.4>0 foralll.3 Coordinates of Vectors of Joint Sets and Sliding Directions for r : (0. Then knowing.3606 0.0r < 0 (e.31).3873 o.2923 0 3zs 0.31) Two-plane sliding. the ( signs in (9.0.2e) Single-plane sliding.i is parallel to planes i and7. i'4>0 foralll.f. When a block is lifting.8660 0.30) and r.

dr : sign (s'fr)& fot I * i (e.3s) For lifting. An example will be worked out.33).866W.A. In this case.A. : (0.500H/) yields -. following the logic of the previous cases. From (9.0.0.)fit (e. . using first vector calculations. The Kinematics and Stability of Removable Blocks Chap. 9 therefore identifying the JP.5WW). i: f. using the data of Table 9. is (1100). yalues given in Table 9. 6t < 0. if S. 6. corre- sponding to r.34). -W) VECTOR SOLUTIOII FOR THE JP GORRESPOIIDI]TG TO A GIVEIT SLIDITTG DIRECTIO]T Lifting. Two cases are considered. 0r : sigt (s'fi)fu for all / (e. Therefore. and (9.3 with r: rz ( From (9.36) Also from (9.3e) and 0t: -sign (St'fi. the JP in the lifting mode under the action of r. subsequently. More succinctly.0. Single-face sliding on plane f.38) dr: -sign (ji'fi. Since from (9.f : ir. and. ) 0 and 0t: -fir if S.29) 3.)ft.30) S.fu ( 0. Using the fi.2. 0. stereographic projection. 4: fu.. accordingly. (e.31). ) 0.866W. for sliding on plane L (i:1). (9. : (0. | . ) 0 for all I # f..'t-i 6t: -fir 0s: fit U=O L=l 0z: -fi2 0t: fit Accordingly. is determined by 4: sieR (Su.3 and ..37) For example. 0r : -sign (r ' fit)fit (e. j (e.32). The joint system for the example is determined in Table 9.7. 0t:ftt 0z: -fi2 0t: fis 6+: fi+ So the JP that slides on plane 1 is (0100).40) . I + i. and r.0. Double-face sliding on planes i and j.

o.4 JPs Corresponding to Each Potential Sliding Mode Sliding Direction r: (0. Recall that . 9. 0.0.9229) when r : (0. (The stereographic procedure for find- ing "the orthographic projection of a vector on a plane" was discussed in Chap- ter 3.Stereographic Projection for the JP Corresponding to a Given Sliding Direction 3O7 Consider. yields 6r:fit 6z : -fi2 6s: fis and 8n: fi+ So the JP that tends to slide on planes 1 and 2 is (0100). Pt. and.500W) and for r : (0.6 shows the projections of the four joint sets previously considered. TABLE 9. Table 9. that iso the intersection inclined less than 90' with l. is one of the two points where circles i andT intersect.866W. 0. Figure 9. . The . -W).ir 1001 1110 . . -W) r: (0.Table 9.3 gives 3rz: (-0.f+ 1100 1 111 3rz 001 I 0000 srr 0001 0110 jr+ 0110 0111 3zt 1101 001 I 3z+ 1000 1001 Ss+ 1110 1101 STEREOGRAPHIC PROJECTION FOR THE JP CORRESPONDING TO A GIVEN STIDING DIREGTION All of the preceding analysis can be performed graphically. O. directions are no longer dip vectors.f. 0111 0100 Sz 1011 1000 .500W).6 when the resultant is due to weight alone I r : (0.4 gives the JPs for all modes of failure both for r: (0.7 shows all sliding directions for the joint sets of Fig.f.f.0.866. 0. sliding on planes I and 2. .In this case.W). is the orthographic projection of f on plane i. and ^i' are the lower-hemisphere inter- sections of P.O5OOW) r 1111 1100 .0. 0. 0. it is the one closest to f.f.866W. Figure 9. .8 shows all sliding directions for the resultant r : (0.3005. correspond to the dip vectors in each joint plane. for example.3. Considering each of fib as given in Table 9. 0..866W.24M.500).) Also. Figure 9.

7 and 9. Lifting along r. (If the resultant force is contained in a plane. Double-face sliding on planes i and 7. 9.34) requires that the JP be on the side of plane i that does not contain . the JP must be outside of the circle for plane 7. t < 0.1. . the JP that contains f has been labeled by a "0.8. The former. (9. Similarly.33) and (9.f. if . When sliding is in plane r.7." Single-face sliding on plane i. that is. from the proposition on single-face sliding. These two conditions describe a unique JP. The Kinematics and Stability of Removable Blocks Chap. the JPs corresponding to each mode have been .) In Figs. Moreover.8. there is no lifting e (JP n P. S. requires that the JP be on the side of plane j that does not contain ^f. This requires that the JP be on the side of plane i that does not contain i.6 Projection of the jont data from Table 9.Ai { 0. The additional conditions are given in (9. One corner of the JP is . The JP that tends to lift is the one that contains l.i.. 9.34).. There are usuallyfour JPs having this corner.) meaning that one side of the JP is the circular arc length that contains Sr.7 and 9 tlll Figure 9. In Figs. is inside the circle for plane j. 6t .

-W). Figure 9.8 All sliding directions and modes when r: ( Figure 9.7 All sliding directions and modes when r : (0. .500W).0.

Com- parison with Table 9. TABLE 9.5 summarizes the comparison. O. 9.1. Fig. -W) Type r : (0. Those that are removable but lack a sliding direction correspond to stable.7 examines the same joint pyramids as in the underground chamber case studied in Chapter 7.1I. . 0. removable blocks in the roof are those whose JPs are entirely outside the reference circle and removable blocks of the floor are those whose JPs are entirely inside the reference circle.8 gives the sliding directions for each removable block. Comparison of Figs.5 Mode of Sliding Mode of under Gravity I Block Sliding with Block Removable Blocks: r: (0. a JP needs to be removable and have a sliding mode. COMPARISON OF REMOVABILITY AIUD MODE ANALYSES A keyblock must be removable and have a sliding mode. (Roman numerals refer to the blctck classification given in Table 4.0. 9. Table 9. The reference circle is the projection of the roof and floor of the chamber. type III blocks.4 shows that the vector and graphical methods give identical results. Recall that only certain JPs have a sliding direction.7.) FI]UDING THE STIDING DIRECTION FOR A GIVEN JP In the preceding section we determined a unique JP given the resultant force r. In particular. 7. Overlaying the two analyses then shortens the list of potential key blocks. Here we treat the inverse problem.866. Removable blocks lacking a sliding mode are stable. and 9.310 The Kinematics and Stability of Removabte Blocks Chap. For example. In the following section we establish a criterion for a removable block to be stable. g identified for both resultant force directions of the previous examples.500)W Type Of the roof tOl1 Single face along IorII Double face along I or II 3z 3rz 1101 Double face along IorII Double face along I or II 3zt jr+ Of the floor 0100 None UI None III 0010 None III Double face along I or II 3zs In order to be a keyblock (I) or potential key block (II). (type III).

r<o (e.43). The proofs are presented in the appendix to this chapter. $ is "a nearest vector of JP with respect to f . For computation. all .i). g) the angle between f and f. Proposition 1.41) Criterion 2. Example: Gomputation of Sliding Direction and Mode By using the propositions about the nearest vectors.-W). Any JP satisfying (9. I .. i> 0.44) is seen on the stereographic projection as a spherical polygon lacking any sliding directions.500).7). (0100). If g is a nearest vector of JP with respect to r. g shall be called "the smallest angle between JP and i.0. (9. d JP for all ij (e. fr. r : (0. 0.43) . then S is the nearest vector of JP with respect to r and r . This is demonstrated by the following example." The following three propositions deter- mine the nearest vectors of JPs for all cases.44) The second criterion permits computation of stable JPs. If 3 is the sliding direction of IP under active resultant r. Griteria for Stable Blocks These propositions generate criteria to judge whether or not a JP ccrre- sponds to a stable block. If there is a vector rt e JP such that (i. then there is one and only one nearest vector of JP wtth respect to t. we can compute the sliding direction and the sliding mode of a given JP. and all . Proposition 2. and (9.0. Fig. when r is (0.8 shows that (0001).. 0. Under active resultant r : (0. and I ..f. (0000). € JP.n1 <90".42) ir + JP for all i (e. .866. (1011).r) 0.f.42). and (0011) satisfy criterion 2.866.500). for example. If there is a vector $ e JP such that(i. based on the same joint system as discussed previously in this chapter. A JP corresponds to a stable block if for uny fr. and (0010) satisfy criterion 2 (Fig.Finding the Sliding Direction for a Given JP The Nearest Vector of a JP with Respect to r Given the active resultant r and a joint pyramid JP. for any vector f e JP we denote by (f . are not contained in JP: i +JP (e. 9. 9. Proposition 3.i. Criterion 1." The angle Q..e.0. A JP corresponds to a stable block if r.f) is less than or equal to the angle between i and anyvector of JP. then $ is the sliding direction (i.

5196 0. Compute the sliding directions using equations (9.2l ^ir 0.t2 Now compute their angles with r.7 lists the sliding directions that TABLE z. calculate which sliding directions are contained in the JP.500) Sliding Angle with r Direction X z (dee) .f r ^ig 111 I 3+. . .2l13 42.8262 -o.38 ^f+ 0.3873 o.f l. Recall that each JP corresponds to the solution set of a system of inequalities. .ir 0. Sliding Directions Contained.333 Szs -0.8) and (9.339 Srr 0. Sz+ Sz 1001 izs. Sz. Now for a given JP.52s6 -0. using (2. . tmo Stz.0500 o.f r. and JP Sliding Mode 0000 srz 3rz 0001 Stable 0010 Stable o0ll Stable 0100 .6415 82.32r il.4233 0.3z* Sz+ l0l0 Tapered 1011 3zs .@ sr+ -0.5587 0. The results are given in the right column of Table 9.ig + r +..1968 27.8949 38.9755 -o. TABLE 9. For convenience the coordinates of all .6.866. jr+ .fzs 1100 f.2r74 52.32s. itz.334 3s+ l110 31.6811 0.26). f) for f: (0. Sts'ize.6 Angle (3.2923 0 75.f.2186 0.31.3890 0.ic .27 3z+ -o. h+ .5973 48.$r r 0111 Sr+ . A sliding direction belongs to a JP if and only if it also satisfies all these inequalities.9563 o.34 3tz o.336 3z -o.312 The Kinematics and Stability of Removable Blocks Chap.6.8977 g .30 jrc 0.9745 -0.9). 9 1.0599 o. are repeated in Table 9. Table 9.2185 42.5755 to.804r 0.3606 0.0.324.1487 0.7 Nearest Vector. St+. .frc r 1101 32s. 0.2097 0101 Tapered 0110 $rs.91t6 0.

A region is called "0" if the sliding mode is lifting when I plots inside the region.Mode and Stability Analysis With Varying Direction for the Active Resultant Force 313 are contained in each JP. parametric studies are performed to determine the most critical direction. If i is a unit vector of any orientation. However. along their line of intersection.7. meaning safe. The symbol i identifies a region for i in which the mode of failure for the JP is sliding on planei.6. This region will be labeled S. The magnitude of a contour at a certain point within a region will indicate the value of the friction angles on the sliding planes in order to produce a net sliding force equal to zero. Another class of problem with variable direction for the actual resultant is posed by analysis of resistance to earthquake forces. One class of problems of this type concerns water forces on a block in the abutment of a dam. different modes of sliding and different magnitudes of the sliding force will be calculated. we will prove that these regions and contours derive from the equilibrium equations. Subsequently. This is the friction angle required for limiting equilibrium when the resultant force acts in the direc- tion corresponding to the point in question. there are other practical problems in which the resultant force does vary in direction. the range of its stereographic projection is the entire projection plane. . was constant. The sliding mode corresponds to the sliding direc- tion that makes the smallest angle with r. since such analyses must be performed without sure knowledge of the direction of the inertia forces. the symbol y establishes the applicable mode as sliding on planes i and j. As the direction of the active resultant shifts. it is possible to construct contours of friction angle required for equilibrium. there will be an additional closed region lacking any sliding mode. problems of block sliding under gravity develop with unchanged orientation for the resultant force. Indeed. We will subdivide the whole projection plane into regions within which f can move without changing the mode of failure. MODE AND STABILITY ANALYSIS WITH VARYING DIREGTION FOR THE ACTIVE RESULTANT FORCE In the preceding discussion. as sliding is impossible when the resultant force plots therein. The "equilibrium regions" will be labeled as follows. Similarly. the magnitude and direction of the water forces on the faces of the block change as the reservoir level fluctuates. Each makes an angle with r that was previously reported in Table 9. we assumed that the direction of the active resultant. Let us now confine our attention to a single joint pyramid. f. The sliding modes are listed for each JP in the right column of Table 9. We will first establish a construction method to bound and contour the equilibrium regions. even if the friction angle is zero. Within each equilibrium region. This simplified the analysis of sliding modes in the different blocks. When the whole projection plane has been subdivided in this w&y.

. for a plane (p. Denote the region corresponding to the JP as "0. The region inside this circle is the projection of the JP. In this case JP is the upper half-space. Consider one particular JP. g Construction of Equilibrium Regions for a Jp Consisting of One Joint Set Consider first a case of sliding with only one set of joints. Figure 9. The locus of points making a fixed angle with a given line is a cone. is upward.1. -f . points out of the JP. 6. points into the JP. Let 0t : (-X..13) as R -----7- | : lzl .6.." The region outside of the JP is then denoted "1. Let fi. drawn as dashed circles. cos a) 6t:fir and frt : -fir Gonstruction of Equilibrium Regions for a Jp Gonsisting of lr Joints (z 2 2) Consider the system of four joint sets listed in Table 9. The radius (r) of the required small circle and the coordinates (C". The steps in the construction of the equilibrium regions are as follows. -Z). unit normal vectors of plane P. Draw the great circle for plane 1. The radius of the great circle for plane 1 and the coordinates of the center for this circle are given by (3. 9. Assume that the friction angle of the joints of set I is f1.-itwill produce a net sliding force equal exactly to zero.: L3" and B : 343o. (ararasat).e. If the active resultant is inclined exactly Qt from the normal to plane P. such that fi. sin a cos B. 1. The method for constructing the equilibrium regions for (arararar) is as follows. with fir).45) Z rt and 'z RY l-r.10 and the JPs are labeled.9 shows a family of such small circles." 3.) with 6..314 The Kinematics and Stability of Removable Blocks Chap. and. : --=- where R is the radius of the reference circle. The stereographic projection of these joints is shown in Fig.A RF u*: (e. making an angle of d. so fir : (sin a sin p. Cr) of its center are given by (3. this projects as a small circle about the normal pointed out of the JP (i.20) and (3.22). rD. and r0.

--a' .r. -_--t -------"/'/' . -/. \-'\. -\-\-.. . --.9 Contours of required friction angle and equilibrium regions for one joint set. 316 .l \ tiffi\\! I I I \t. -----r-// \-\-'--------. .10 Lower-focal-point stereographic projection of joint data of Table 9./' . ta -at' \r.1.rr.------.*. . '. \ t/l/ tt\ I \\\iiruT. 1111 Figure 9. .r.\\:__:::_-- \\--= ---.. ---. i I / / / _\ .ra{r/=zo" -t- Figure 9.r' ---./' /\\ \ I! t'/ --.

Project the JP by the usual stereographic projection procedure and label each of its sides as P.i is the stereographic projection of i. 9. equatipn (8. This can be accomplishec as follows.316 The Kinematics and Stability of Removable Blocks Chap.11. and.with the values of the indexes i andT determined by the common sides.r. pn. We can compute the edge vectors and their coordinates using an 'oedge matrix. C." as estab- lished in Chapter 8. (Fig.) of normal vectors Qi.) and construct great- circle segments connectin g w. compute the projections (vr.I2) established a function I(a..p. e . r. "6Dr. . crn.. and. wzwl Figure 9. e . whereltt : fi. On the stereographic piojec- tion. 2. For JP (arararan).11). the clockwise order is p. 9 1.). e . 3. pr.1. -f . Fig.r.o: lrn (tn the above. For example. 9. Then . : -1rr. the four corners of JP 1100 are labeled. The corners are. Irn. : -frr. and c'3. with friction angles as given in ttrat table. comparing with the calculated edge vectors.. establish the clockwise order of the sides. X fir. for JP 1100.. Compute the edge vector of the JP and find its projection. Cu.o: Irn. crn. C. crz. The joint data of Table 9.4(a) as -lrr.1 arc the same as used in Chapter 8 and the edge vectors of JP 1100 are given in Table g.. e .ll Lower-focal-point stereographic projection of equilibrium regions for JP I100 of the joint system of Table 9. in clockwise order. frn.

. This region is denoted..r. the regions are as follows: lifting occurs in the polygon between Crr. x i.11 for JP 1100. and C. and 13. proceeding in the order of the sides of the JP. Plot w.o ar" in plane i.. mode 0." Double-face sliding on planes i and 7 along direction ^i. and C1r. Lftd C.)rD.. This will be constructed as a dashed circle.i. 9. Single-plane sliding along plane i in direction .a). * sin (6)e . even with zero friction on the joints. If rD. Crn..2) Then its projection is wi : (Xo. : fir.)fi. It is important to observe that some modes of double-face sliding cannot occur. For JP 1100.. and e . Using the same approach. wrCrr.34. and wtwr. draw great-circle segments connecting w. and C. (9'46) For JP Il00 (ataztja+: 1100)... in this example. This is termed "mode i.. is the normal to plane i and both e . D. 9. 5.. Double-face sliding occurs in modes 12. wsCt+'. Y.I1..f. w. Lifting.Crn. and i.Mode and Stability Analysis With Varying Direction for the Active Resultant Force 317 fit: -I(a.) Using (3. then woto r'3' and' flnally.S..r. fi. wn plots off the drawing. This great circle will pass through points lr'. wrwzi wzla+i wtws. modes 23 and L4 do not occur. and Cy for the projection circle of the plane whose normal vector is rD. is perpendicular to both e .wn." No sliding can occur.lin the plane of fi.13). and e . to wr. to Cs. is upward. This is termed "mode ij. Erase the portion of the circle that is greater than 180". as shown in Fig. Yo) where the values of -tr' and y0 are giYen by (3. Because tD.o. and C.fine t. wsCrt. wt. draw the great-circle segments connectin g wt and wr. Single-face slidingoccurs on all four faces. In the example of Fig. is defined by the region inside the JP. Finally. and fi) 4 : Suppose that -fr* frr: (X. draw it clockwise.. draw this circular segment counterclockwise from w.. is delimited by the spherical triangle whose corners are Ct. efld wr. compute r.. Draw the circle of zero sliding force for modes i corresponding to a given value of the friction angle f.o.47) .. tu : cos (t'. the great-circle segments constructedaccord- ing to this procedure a. ftt : -fir. C11. For JP 1100. (e.. when the resultant plots inside the spherical polygon whose corners lta tuiT w11 wk. connect wrto wr. 4. fi r : fit. otherwise. fr. is delimited by the spherical triangle whose corners are wi. Assume that the corners of region i ate w.thenwrto wn. Identify the sliding mode of each region established in the preceding construction. C*. x i.24.o.te w1C1zi wzCtzi wzCz+i w+Cz+i wtCs. (In this figure.

u is inclined f. is in the great circle wtC.318 The Kinematics and Stability of Removable Blocks Chap. are thlir stereographic projections. at a glance. great circle tz+t+z is drawn in region 24. and is therefore the locus of zero net sliding force bn plane i. The points tq Lte plotted using equations (3. 6. from wr. und tr. These circles have validity only within the appropriate regions and therefore are erased from all points outside the appropriate equilibrium regions. from wr. and r1s. The dashed circles in regions l. the dashed locus connects tu and t2n.z. The friction angles for the four joint sets stated in that table were used in the constructions.48) Project the plane perpendicular to fi.l w€r€ calculated from (9. the point /r.10 and Table 9. draw the circle of zero net sliding force for modes 1.and 3 are shown in Fig.: cos (rf. is in the great circle wtCtt. each of which are fi.)rO. Figure 9..r.e .. the dashed line connects points /.20) and (3. both f. with wr. 9.o. In the i.z) (e. g The vector i.13 examines the equilibrium regions for JP 1101.. In region 2. vector trk given by t.13). what friction angles are required to achieve equilibrium corresponding to any posi- tion of the active resultant. Crr. great-circle seg- ment t+sttr is drawn as a dashed line in region 34.. the great circle segments tz'trzare drawn with a dashed line in region 12. In region 1. we calculated the normal to the plane repreiented by this great circle. 9.t. fi. the contours can apply only if 6t : f1' This type of diagram can be useful because it shows. Figure 9.7'. This great circle will intersect circle wtCtt at t11and will intersect circle wtCu at t11. To construct it. The vectors f .. from r0.47) and t.12 shows additional contours of zero net sliding force for 10o increments in friction angle. Now construct a small circle from t11to lr*.. That normal.11. both of which make an angle of 6.7 are defined by . * sin (6)e . and segment tsrttt runs through region 13.1 und f.. And in region 3. In Fig. Similarly.7 regions.1. in the plane fr. 9. is computed by ft. and wr.:##:(n. The required locus is a great circle connecting t.r. in plane fr.r.11 (their continuations in region 4 are offthe drawing).Similarly. Examples We have examined in detail the example posed by construction of the equilibrium regions fcrr JP 1100. This represents the cone of constant angle {r with the normal fi. with the joint . is inclined f. using equations (3. the small circle runs from 131 to tro. The corners of regions f.4) and the small circle radii and centers are located using equations (3. from fi.. with the joint system of Fig. Finally.22).e . and t 1. The point r.

--" Figure9.l3 Equilibrium regions for JP 1101.i:'. | :' I 41ii. 4 t.1.12 Equilibrium regions and contours of required friction for JP 1100.rt\r \ i X\. with friction angles as given in Table 9.| \ i 11. Figure9.ry ?:. i'i 1 . . assuming equal friction angle on each plane.'\\i}.

e .14 Equilibrium regions and contours of required friction for JP l l0l.14 shows the contours of zero net sliding force corresponding to a range in friction angles in 10' increments.8 Dip. 9. JP 1101 has onlythree corners and three sides./ -- 'll. 3. The position of 1101 relative to the other joint pyramids is shown on Fig../ . and 4. 34. and 23./ . The sliding regions that exist for JP 1101 are 2. Even though there are four joint sets..10. 70" r 60o /qo" 3 I \ \ \ \ '\\ r\'r \\\\ \ Vrt /..15. This Jp has twoedges./ /.1. Figure 9.----\r / / .320 The Kinematics and Stability of Removable Blocks Chap.' _--- /.' /- 4. a Dip Direction.. 9. the particular JP has only three faces.8.jt+1'- 23 0= 10o Figure 9.///. Consider the joint data given in Table 9./ . 9 system and friction angles given in Table 9./ . The joint pyramid (arat): (10) is shadedin Fig. Special Gase: Equilibrium Regions for a Jp with Only Two Sets of Joints A JP with two sides presents an important special case of the previous construction procedures. p Friction Angle Joint Set (dee) (dee) (dee) I 68 243 20 2 45 280 40 ./ .rand TABLE 9./ \\ /// / . and 24.

8. from rD1 toward -e . representing single-face sliding along plane 1 in direction . from r0. with r0r.L. similarly.. . and ur.. is in the region between great circle 2 and great circle C12w2C'1r. and *dr . Mode 0 is in the projection of the JP. Mode 1.. inclined {. The regions are as follows: . and -e . arc both perpendicular to e . Vectors fi1and fi. Mode l2 is in the spherical triangle between great-circle segments Ctzwt. (and e . wittr i. . Ctzwz.". is in the region between great circle l and great circle CrrwrC'1r. and. Point /r.wr. It is contiguous to the JP.r). . The vectors 1 and li/2 -e . respectively. e t2 and makes an angle f .i.r. respectively.Mode and Stability Analysis With Varying Direction for the Active Resultant Force Figure 9. is the projection of vector ir. .". are C* and Cl2. andis in the plane of r0.r. whose projectionsplanes rD are outward normals to L and 2.r.Point 1. which makes an angle 6. this region is contiguous to the JP. representing single-face sliding along plane 2 in direction . and t'1ris inclined $. is the projection of t tz.on e . Mode 2... i.15 Equilibrium regions for JP 10 in the two-joint system defined in Table 9. Mode -12is in the region remaining that is between great-circle seg- .rinthe plane fr"e . toward -e .ir. is in the plane of r0. It represents sliding on planes 1 and 2 in direc- li. Point t'12is the projection of vector t\.

then. Denote by . ----iYt-.i iI !-/--. formally. \ i \ \t.2\\Lllr\jiJlii / I ii 0=10" d= 10" Figure 9. 4oo \ \ \i\o' -\r \ i. Region 0.. then rc Jp.=l\--=__::. --r-rt. the connections between the equi- librium equations and the constructions above.-a\i\t).""\ZZ *i.\s i i!l/! Iii /'--- _.. occurs along the great circle wrwz between regions !2 and -12. and wrwz.r.r'i'll)N i /i /.wA] --\------.othe projection of the active resultant r.V -\i. \ttrr'.. 9.\1.\----y{_'Y))h\i '*i\ Ii ir. which is safe from sliding with zero joint friction.r. From the proposition on lifting. -. ' Mode ^S. the sliding mode is lifting. \ \ \i\ 1++zz>>_-'rr. lf r" is in region 0.--\ \. l. : Qz. -tZ.r^i..16 Equilibrium regions and contours of required friction angle for JP 10 assuming equal friction angle on each plane. In the regions of modes 12 and apply only if\. "\'\r -----\rY -------__ I 30o---_. 322 The Kinematics and Stability of Removable Blocks Chap. -i.. This mode represents sliding on pranes 1 and 2 in direction r._____ __ :_.8 and Fig. Proofs of the Validity of the Gonstructions for Equilibrium Regions We now proceed to show... these contours 600 5oo 40" 7o"l----\\\\\ \ \\ i 5oo i A . g ments c'rzwt. Figure 9.----.'f i --u1.-----i^t:l@i i i'.\ .'.. c\zwz.15 shows the equilibrium regions for the joint data of Table 9. .16 shows the contour values for these regions corresponding to 10" increments of friction angle.

.s. Regioni.50) becomes r : -{+ cos 0j (cos d. . tD. -73. The friction angle of P. * sin f..1. into (9..T)0suchthat r:Ntfit*Tit and r * l[rdr .: 3tl. and ir at an angle f. from r0..73. and C*.-N' tanf..1. (fi. and ro belongs to region fl Assume that rolies in region l.f. 0.d. (e. /Yi tan frsr .50) In the case of limiting equilibriun. F: 0 and (9. i/. .3). L C.) then i: cos 6. the projection of I to plane P1. fi..o. * sin {. as axis and angle {.Tft: g (e.53) which is the limit equilibrium equation [equation (4) of the appendix to this .* Nfit . w.. [given as equation (1) in the appendix to this chapterl. The projection ro of f is in the small circle with rD. (e.Equation (9. From (9. Region y.(r}. Cii. and T>0 such that r:Nrrit*Ntfit*TC. Assume that r" lies in the region U. Because region y is the stereographic projection of a pyramid with three edge vectors.A.. which has cornets rr121rrit and C.51) which means that i lies between lil. having corners wr.1. F3. i. When I is between fit and ir. L to. and r belongs to this region.49) gives rf N. 6.. l/j>0.52) becomes r I N. . there are values of Nd>0.: 3. there are values of Nr ) 0 and. : g giving r : f tan di3r) + Fi.:g (9.. Substituting this value of -2.i.i..FS. and Since wt L C.t (e..s2) Because e ..Mode and Stability Analysis With Varying Direction for the Active Resultant Force 323 2.f.4e) This is the limit equilibrium condition for sliding along . L fi. lies between Cil and Ctk.> o so e . 3.ltrd r lies in this pyramid... x i) x rfi.e.

.54) becomes r : i[.t (e. tanQ.ff.3. N.. sliding on one face. then i.3.f.13). so the sliding mode is double-face sliding along ... t..f. .1 . Qf .S. and f...22).t * ffr > 0. sliding parallel to . the angle between f and f is greater than or equal to 90'.r0r*sini F \. The Kinettratics and Stability of Removable Blocks Chap.1[.S.. such that r : N. & r..) + Nrli.J Substituting this expression for -fitt into (9.. from the proposition of (9.FS.S.r. < 0 Then g+JP This is a contradiction. we have r{ N. Then ro lies in the great circle connecting t. There are lr. Assume that ro lies in region. Region ^S..>0. Then there is no direction in the region that can satisfy the propositions for lifting. Nj>0.i..0.s4) In the case of limit equilibriun.. N.3.(rli * tan i.55) shows that r lies between f tr and f The projection of I is in the .. Equation (9.- cos yr. .55) ffitt. From (9...) > 0 8'fr > 0 giving 8'6.. (e. x ir. F: 0 and (9. in the reglon ij. So for any g € JP. such that g'(l[lf/.fii * Ntfrt * Noti. The block cannot slide even if the friction angle of each face is zero.E > 0 There is a vector r0. . and.o+ . tan Q.FSu: 0 or r : tr/.1 .i) * lfi(rAi * tan 6$. Denote the corners of region as ^S W17W1tWkr.) r:-*=-(cosf.r*I' r:: d?.53).ffi<"os 6fit * sin 6t3') Finally. The normal vector of this plane is f. lrk>0. > 0 (9..t * tan 0. or sliding on two faces. t1. + N jOr .S.. N j tan . requires that -73tt : *.lfr tanQ.(fi. fi' s'i) t . projection of the plane of f .. * tan fi. 9 chapterl of double-face sliding along . 4.56) For any vector SeJP if the angle between g and I is less than 90".) * Ff.

(0. t * N.(0.6.5) becomes r*i[iOr-T. ! Np1. (9. . The normal components of the reaction forces on the sliding planes are therefore : N.*Nfit-73:0. 9 Appendix APPENDIX PROOFS OF PROPOSITIONS 1.f(.-6o so the above can be written o_(fi.nrr > 0.lri > 0 1**. all of the joint planes except those of set i and 7 will open.T(6.i:0. on the right. with N. (6.(0.lfj>0. .i x dr).(6. and then the dot product with (dr x Dr) gives (r x d). x 6) .x r) x 0. Nd > 0 (3) Then since N. end.Chap.N.i then (6rxr)xd':TS (2) and "^_(0rxr)xd. r ( 0..A..0. (6' x dr) : 0 (5) . * 6t). x 6t) + N. PROOFS OF PROPOSITIONS ON SINGLE- AND DOUBLE-FACED STIDING Single-Face Sliding Since . .. The equilibrium equation. on the left and then 6. becomes rf. f>0 (1) Taking the vector cross product of equation (1) with 6.. and P1. 0.0') Since . x ^i) x dr : S(6t._o " -l(ri x r)T--"' Taking the dot product of d.i lies only within plane P. accordingly. with equation (1) gives 0t. x ^i) x 6t : O (6.5).B is removable. If/>0.xr)xfi.6. Double-Face Sliding Since .($. the equilibrium equation (9. ) 0.>0. . . N.i):. the normal component of the reaction forces on the block reduce to N.l(6.i is contained in P.x r)T- ftt: 6t ot fi.0. . . T>O (4) Taking first the cross product of (a) with 0.) : 0. ) 0.

Proof.326 The Kinematics and Stability of Removable Blocks Chap.n) < 90" then there is one and only one nearest vector of JP with respect to f . s" : (i' x Sign ((fi.x fi. (r x 6r). r > 0. and then taking the dot product with (d.Z) Er : (At. * 0.) gives . <0 (g) Taking the dot product of (a) with 3 gives r. ( 0 So i.) : (0t x 6r). * N. x 6. PROPOSITIONS OIU THE NEAREST VECTOR Proposition 1.).6..(0. s. JP such that C. C2) Let Etz:#+€h (see Fig.u.f. when (ff. 9 Since ^f is parallel to (6. dr) : -(r * 6t) x 6t'6': (0t x r) x 6''6' so (5) reduces to [(6r x r) x dr]. x 6. r) lfu x \t)r. [(0txr)xd/. (10) requires s:fitx fiiand when (ft. < .{0 or I@t x r) x n).0. & >0 (6) Therefore.(!. then i'8r: i'8r) O Denote i:(X. Also. (10) requires s: -fir * fr. fi1: 2.i)-0 (9) Since T > 0 for a key block or potential key block. x ft) .f-f(^f..di(O (7) Similarly. 82.. x 6r).d. multiplying (4) by 0... If there is a vector fi . r. Bb C) 8z : (A2.y.r)0 (10) Since the sliding direction is parallel to fi4 x fi1. x fi) .(0. x 61) : 0.6.(r x : 6t x (r x 0). the last term : 0.). Suppose that there are two nearest vectors gr and fr. therefore.

Ifi : jr. the sum of the lengths of the two sides S. 1.17 In the force triangle of gr . this is contradiction. then.i lgr gzl 2 or 9rz. 52.) (12) . and $. we have S. From Fig. the third side's length is less than 2.i: . and f . then ^i. is the orthographic projection of f on plane P. 9. Figure 9. Proof.r. 18'*8rl<18'l*l9rl:2 Err. If . 2. 3r.18 it can be seen that OB: OAcosu O. 3. . r so (i.f is the nearest vector of JP to f . If 3: i. or .f is the nearest vector of JP to f.. rt) (1 1) where fr is any vector of plane P.i : L and . Proposition 2.8') $.. i >(8' -r-9)'i :gr.r ) Er.i.) cos G.. fi) < cos (f. equals 2.f. and 8r * Pr.€'!-t 8. .|'r) < (f. .i is either i.i is the sliding direction of JP under active resultant r. From (11) it can be seen that cos (f . Therefore. fr) : cos (1..Bcos p:mcosdcosB -N: a/a ?: 'oA'L: COS COS LCOS B cos (f.i > 0. and $rare not the nearest vectors of JP to i.

and Pl such that 6.) n Pr) lU(6. is the nearest vector of plane P. from (11): ( (13) and cos (i.ii) cos (^fi.) n 4). the nearest direction is unique and equals g. u for any rt e P.1o. 3.) n u(6).ir) cos (5r.) n Pr : Pt For any h e (U(0.l u n ^U(6r). . If . pointing into the JP. P.c . and d. point into the JP.r) : cos (f . € JP. SinceJP U(6. consider u@.. Prl : (U@t) n PJ u (U( As before. therefore. the nearest vector of the half-space u(d)with respect to I must therefore be on the boundary plane. g is the nearest direction of half-space U(a) to f. meaningthat^f. when i-6. letfl. iri) (14) .) The last step follows from ^ U(0) A Pt : Pt and U(0. from (9) the normal vector of f.f. its boundary is the union of two half-planes: (U(6) . i.9] Figure 9.s) is smaller than e.)and. andfr#3i then (l. U(0)) (P. By (12)..328 The Kinematics and Stability of Removable Blocks Chap.f > 0. fi) : cos cos (f.) : lu(o) U(6) n p. By Proposition 1. The dot product of equation (1) with S shows that i.l. . i. i + u(flt).i : . is also the nearest direction of the JP to l.3:f>0 and (1. f) < 90'. Denote by 6t and 6t the normal vectors of P.

. If f is the nearest vector of JP to the active resultant f and $. . so Si + U@. 2.i) ) cos (f . 9. i: $and I : j is the sliding direction. then $ is the sliding direction i E -: 3. is the nearest vectorof U(6r) ) P. lf $ is in the boundary of the JP and belongs to plane r but not to any other plane of the JP.) to l. 9. $ is unique..) ) 4^is shaded and ^i.i.^Consider another vector ft in plane P. 9 Appendix For double-face sliding. s.) becomes smaller.f. and JP c U(6. This is impossible because $ is the nearest vector of JP n P. i.Chap..-i 10..f.-Si <0. Therefore. We denote any vector in this half-plane by h. ^ Proposition 3. In Fig. when the angle (fr..19 Then. i. (9.. In partic- ular. Similarly.i.i) Comparing (13) and (14) gives us cos (f. Because i+U(6. If $is notin the boundary of the JP..34) gives $. Praof.19) u(0i) n Pj Figure 9. Therefore. the angle e.. $ : Sr. is the nearest vector of U(6. . iii € JP. then 6.).)^U@j) the nearest vector of U(6. Therefore.h) ( (ii. meaning that g is the sliding direction and the sliding mode is single-face sliding along J. 1. fr) (see Fig.19 the region U(0.When ft moves from $ toward . theng e JP. that is. is outside of it.) A Pi.) ) U(A) of i. 4 U(6. Because g-i > 0.y is the nearest vector of JP to f.) U(A) shows that ^i.) ) U@ j) to I is in its boundary.i > 0. G. ^f. to r. From the figure. is the nearest vector of U(6') o Pt to i.b calculated by (11) also becomes smaller. and suppose that $ # 3. then.. the mode is lifting.irr is the nearest vector of (U(6J n Pr) U (U(0) n P.

we have 6r. JP n P. then I : 3. Fig.s.r.r.33) and (9.0. from (11). such that 6. (i. the boundary of which is j' and . < 0 Then by (9. \\\\ \\ 3'. so 6r. n P7.6rand. 9.20 When fi e Prmoves from E : S.f.20 Similarly. are the normal vectors of P. fi) also becomes smallel. If $ e P. < 0 as shown in Fig.f.20 shows that (J. I The Kinematics and Stability of Removable Blocks Chap.# u(o) n P. fr.) ) (i. i. since $ is the nearest vector of JP A Pr. 9. \\.. and 0r point into JP.As assumed before.1. a. becomes smaller.o as shown in Fig. fi\ Figure 9.34).. 9.f. .20. s. and.. (i.J. is an angle in P. . Because g is the nearest vector to f of JP n P. 9 1 Pp respectively. JP has the mode of double-face sliding along.ttoward i.r..) Also.

H.)." Bulletin of the Association af Engineering Geologisfs. p. and N. O. Opnc^q.S. Moncsr.-Y. Yol. A." Geotechnique. CHIN.238. Proc. "Shotcrete in hard rock tunneling. I9th U." Proc. Bentou. pp. N. GoonuaN (1983). LaNwry. and A. Btsuot. A.. Bnrrxr.. pp. and L. Colo. 3. A.Referoncos AnNoto. and N... Colorado School of Mines. BRocH. L. Symposium on Rock Mechanics (Association of Engineering Geologists and Texas A&M University. "Case histories of three tunnel support failures. Wrr. Paris). T. 8. R.W. pp. Vol." Subsurface Space." Ph." Proc. pp. A. G." Bulletin of the Association of Engineering Geologisls. Paris (Dunod. (1977). and R. 265-300. College Station). "Contribution d l'6tude de la stabilitd des appuis de barrages- 6tude g6otechnique de la roche de Malpasset. 10.S.rsrnnrv (1960). pp. D. Barcnrn..D.259-266. Symposium on Rock Mechanics. (1966). pp. pp. No. 3. ErNsrsru (1977). E.S. No.ano (1980). 241-264. G. Reno. B. 5G65. "Sampling for joint persistence.rrNBy (1978). Symltosium an Rock Mechanics (ASCE.S. Vol. "Stability coefficients for earth slopes. "Prediction of support requirements for hard rock excavations using keyblock theory and joint statistics. P. 5Cl-8. Golden. IX. 24th U. Symposium on Rock Mechanics. (Johnson Publishing Company. IX.q. Ecole Polytechnique. 18th U. "Storing water is safe and cheap.sou (1972). New York). E. 8. Golden. (1972). and H. TflvnEs. California Aqueduct. Brsro. G. L. "Statistical descriptions of rock joints and sampling. N. 331 . 129-150.. thesis. BnEcuEn." Proc. I6th U. University of Nevada. 557-576. BnnNarx. Colo. J.L. Mackay School of Mines.

pp. International Symposium on Rock Fractures. J. Mlnrrr (1962). Vorcrtn. pp. "A computer model for simulating progressive. and J. "Kemano tunnel. specialty Conference on Rock Engineering for Foundations and Slopes. 16th U. International Symposium on Rock Mechanics Related to Dam Foundations. P." Proc.. large scale move- ments in blocky rock systems. 347-373.. Gooontex. R. P. ASCE. Symposium on Rock Mechanics (ASCE. w. A. Vol. F. 567. and J. Bnav (1977). E." Proc. "Toppling of rock slopes. New york). Prentice-Hall. Bergamo. Minn. GoonulN. Vol.E. P. GoopuaN. pp... Introduction to Rock Mechanics. "Numerical analysis of topping failures in jointed rock. p. "Fractical rock foundation design for dams. operation and maintenance. West Publishing Co. Department of Civil Engineering. of Mining and Metallurgy. 8. Goonuarq. (Inst. 45. N. Jotttt. Hunsox. and W. Fuvacnru. (1976). M. HrrrrNcrn. ASCE. W. pp. D. Underground Excavations in Rock (Inst. 94. Symposium on Rock Mechanics (AIME/SME." in Underground Rock Chambers (ASCE.S. (1978). 527 pages. New York). G. I6th U.S. University of California. Engle- wood Cliffs. LoNon. (1977). Lowor. Wiley. Bnl." Proc.. BnowN (1980).37. and E. 23rd U..2nd ed.J. Boulder. E." Proc. llonx.D. Pnrnsr (1979)." Ph. "Discontinuities and rock mass geornetry" International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. New york. Cuunllr. R. and S. pp. SM2. No. . P. p.J." proc. pp. u.o' Annales des ponts et chaussdes. jointed rock using the keyblock principle. "Calculation of support for hard. New York). Vector and Tensor Calculus. New York. Bovu (1982)." Proc. M. (1968). Tanorru (1977). and C. 16. Dover. J. Symposium on Rock Mechanics (ASCE. LBny. 115-138. E." Proc. London).lrnnunsr (1977). pp. (1971). and Gerq-nuA SHI (1982). Hnr. No 7. Colo.. Rock Slope Engineering. J. of Mining and Metallurgy. Rio de Janeiro (reprinted by ISMES. Vol. T. "Graphical analysis of slopes in jointed rock. (1965). M.. GoopnaeN." The Engineering Journal. Nancy. New York). and D. 109).8.. (1980). "Rock engineering for underground caverns. (1953)." International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. 133-137. Jn. 883-898.497-526.E. w. W. Methods of Geological Engineering in Discontinuous Rocks. "Geology and rock slope stability-appli- cation of the key block concept for rock slopes. "Describing the size of discontinuities.. R.v (1976). CnuorN. France (International Society for Rock Mechanics). cononrc." Journal of the Soil Mechanics and Foundations Division. pp..E. New York).. Yol. 14. London) 402 pages. St. Dnnne (1971). "Geomechanical models-notes on the state of the art.. Publ.A.qoBnn. "LJne m6thode d'analyse ir trois dimensions de la stabilitd d'une rive rocheuse.S. J. Grm-nua Snr.E. 3rd International Conference on Stability in Surface Mining (AIME/SME. and B. CuNnrtr. Vol. pp. (1978). 39-86. thesis. IfAv. R. (1952). Fundamentals of Engineering Drawing. Honr E. "Computerized design of rock slopes using interactive graphics for the input and output of geometrical data. E.2. Paul.2Ol-234. D. R. 5-14.339-362. Luz. T. K. I 332 Rererences coor.rDRoN. and J.

Oslo. (1978). Edward Arnold. A.. 17. S." International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. Cambridge.n (1982). E. "Methods to. McGraw-Hill." Inter' national Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. and M.HoN. W. GoopuaN (1981). "The stereographic method of stability analysis of rock masses. PHrrrns. W." International fournal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. (1967). Advanced Engineering Mathematics. A. P. C.D." Proc. Durban. 125-143.260_271. 183-197. 181-190. pp. Snr. No. Vol. and J. "Influence of the shear strength of the joints on the design of prestressed anchors to stabilize a rock slope. Paper 4-11. 1. "The analysis of the stability of general slip surfaces. 15.63-66. K. No. F. GeN-nu r (1977). (1974). "Design of rock slopes against sliding on pre-existing surface. Jrnuqrrvcs (1975). 22nd U. krEsr. Krnc. and R.o'Ph. (1960). "Recent developments in the interpretation of data from joint surveys in rock masses. 3rd Congress of the International Society for Rock Mechanics.C. 3. F. Denver.. thesis. E. HunsoN (1981). pp.D.. Massachusetts Institute of Technology. l. Colo. Vol. S. pp. WArrrs. pp. Washington. Vol.. South Africa. Jn. Vol. pp. C. "Estimation of discontinuity spacing and trace length using scanline surveys.. pp.References 333 McM.3rd ed. Voncnrn. and J. Huosox (1970. Vol. WennunroN. Wytrc.). J. 'oA new concept for support of under- ground and surface excavations in discontinuous rocks based on a keystone princi- ple. R. Rock Mechanics and Engineering Geology. D. pp. O. Vol. SHr. Snr. (1981)." Scientia Sinica. and V. Wrrrrn." International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. 290-296." Geotechnique. GuN-nu. The Use of Stereographic Projection in Structural Geology. II. "Estimating the mean length of discontinuity traces." Proc." Inter- national Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts. 13.221-228. 135-148. 803-808.. R. pp. 18. "Discontinuity spacings in rock. E. New York. and J. IIB. MoncnNsrERN. N. "A stereological interpretation of joint trace data. J.. Vol. Knnnrcu. London. Geotechnical Conference. p.. E. pp. Wrrrrr." Proc. 17. "A geometric method of stability analysis of discontinuous rocks. Symposium on Rock Mechanics. 311.q. 18. (1980)..analyze the stability of a rock slope with and without additional loading" (in German). (1965). Springer- Verlag.79*93. "An interactive graphics-based analysis of the support require- ments of excavations in jointed rock masses. No." Scientia Sinica. p.S." Proc. SrnrFnN. 2nd ed. D. (1971). XXV. pp. XX. University of Minnesota. Vol. Mass. (1980).D.. B. (National Academy of Sciences. M. P. 6th Regional Conference for Africa on Soil Mechanics and Foundation Engineering. P. Pnrcn (1965). PntnsT. pp. H. PAnr. . "Discontinuity spacings in a crystalline rock. Grx-nua. Suppl. 17-26. M. Vol. S. 52. Yienna.

82 Block types by vectors. 29 with curved faces. N. 125 Bamaix.41. 295 Block codes types. 95 numbers of different classes in surficial Angular relationships for an empty excavations. l15 Angles between lines and planes defined. 8 Blocks Assumptions. 98 Barton.49 joint blocks. 162 excavation.247 faces.98. 100 for omitted joint sets. 107 BP (see Block pyramid) 334 . Angles between planes 139 stereographic projection. 186 intersection. 250 Arnold. in a surface Bending.Indox A Block pyrarnid. 130 volume. determination using vector analysis. 3l-38 without repeated faces. 129 united. 19 with repeated joint sets. l2l ' with repeated faces. B 28-40. 8 key blocks.. 4l by stereographic projection. 290 numbers of different classes in tunnels. A. 9-11 corner coordinates. 187 Blasting. J. 105 computed using vectors 4l-43. 40.237 stable..49 numbers of different classes ofjoint blocks.. 125.

4 Complex blocks (see United blocks) at the surface. 157.. 49 EP (see Excavation pyramid) Intersections of underground excavations. 4-9. 335 Brekke.225. L. 3 types. 25 stereographic projection. 194 Discontinuities. 3 Chan.7. 113 underground openings. 173 tunnels.157 Concave slopes.. 5 Equal angle net (see Stereonets) Buckling. 2O3 Complementary set. 103 Engineering judgment. 25 for tunnels. l7l Cook. l5 lntersection line. 13 E Edges of a portal. theorem.8. 3 Half space Distinct (or discrete) element analysis. 232 . 28. 82. 6. 218. 203 Convex slopes. 2 description by stereographic projection. 240 Contacts. 4 defined. 15 is a convex block. 235 Finite element analysis. 6. 147 Forces Crack growth.Y. 14 vector operations involving. 246 F Corners of chamber intersections. 22'7 Fissures.244 Foundations (see Stability analysis) Cylinder. 247 defined. 54 Curved faces. M. 172. 80 excavations. 30 Hittinger. 296 Cundall. 5 purposes. 8 Coordinate axes.229 H choice for a slope. 194 proof. 284 Finiteness. 22'l lnfinite block. 3 Cousins. T'. 101 of an underground excavation. 7 Equilibrium regions. 80 slope angle. 75 description by vectors. 14. 2O8 stresses. 8 Excavations Caving. 85 Direction choice for a chamber. 3 Convex block. 219.9. 105 Castaic dam. 241 D G Design Geological maps (see Trace maps) locations for reinforcement in undersround Great circle. 162 acting on blocks. 320 Excavation pyramid c for curved blocks. 164 modes of failure.. 43-45. 220 center of projection. 190 stereographic projection is a true circle- slope direction. 254 of an underground chamber. 92 Dip and dip direction of a plane. 67-70. 282 I of a pyramid. J.. 113 Don Pedro spillway. 157-62 Clay. 1l-14 of a portal. 313 for a JP with only 2 joints.

84. 88 Joint spacing. 20. 205 unstable vs. 147 Pillars between underground chambers. 16. 16. 75-78. l3l toppling. 205 Limit equilibrium analysis. l& Kinematic conditions for lifting and sliding. 236 in dam foundation. 128.55. 299 types in three dimensions.25 Orthographic projection. I Microfaults.32O Joint blocks. 82 Kemano tunnel. 251 in the walls of an underground excavation. 6-8. 164 Physical models. 135 Joint properties extent. 19 versus joint blocks. 125 single face sliding. l0 M Rock discontinuous.336 lndex J Modes of failure lifting. 45 in the roof. 162. 298. 161 types in two dimensions. 311 s Models numerical. 100 in inside edges of underground excavations.55. 173 . 18 applied to concave slopes. 2 Mines Roof of an underground gallery.296 of a tunnel. 3 JP (see Joint pyramid) Orientation choice for excavations. 3ll. 298 with repeated faces..48. 98. 158. 127 two face sliding. 17 in a surface excavation. 47. 57 -60 of a vector on a plane in the stereographic K projection.54. 310 example. 8 Kerckhoff II underground power house. 45. 139 Multiple blocks. 18 cousins. sliding direction when JP is given 326 by stereographic projection. 2l terms to describe. 150 Joint trace map. 83. stable. 204 Rigidity of blocks. 295 Removable blocks. 150 Joint pyramid Nearest vector of a JP with respect to r. l0 N spacing. 206 L in corners of underground excavations. 4 Mode analysis. R 304 Removable area of a tunnel. 158. 253 Kinematics. l2l physical. 281 P Key blocks analysis. P. 81. 3 hard versus soft. 98 Potential key block. 46. 16l shapes. 236 Lifting mode. 17. ll-17 Shi's theorem. 152 o Joints. 21 I in pillars.21 Portals.295 in underground openings. 99. 83. 279-86 defined. 17. 159. 168 Londe. 24.

295 direction. 53 of opposite to a vector. 152 defined. 1 l0 determining block corners. 54. 38-40 of vectors. 29.24l Statistics. 78 of block pyramid.2I1 Surface excavations. 5 cylinder. 157 applied to portals. 63 equation of a line. 67 compared with stereographic projection of a plane. 238 U of a cone. l0 portals. 303 Tapered blocks. 61-64 Underground chambers equilibrium regions. 91. 47. 48 64 faces of a block. 196 . 287 SP (see Space pyramid) Toppling. 240 Squeezing rock. 54. l2O. 122 stereographic projection is a true circle. 25 upper versus lower hemisphere projections. 105. 283-86 design considerations.74.272 Spacing.lndex 337 applied to nonconvex blocks in undergound Strike and dip. 252. 52. 74 half spaces. given angle. 194 maximum removable area. 138 joint pyramid. 168 single face sliding. 'll. 165 Sliding direction effect of slope curvature. 71 description. 27. 161 Space pyramid Trace maps. 126 of sliding direction. 208.295 face. l2l finiteness theorem. 301 Theorems Slope design finiteness. 48 Sliding force double face sliding. 93 tunnel axis. 314-20 example of key block analysis.24l Stable blocks. 51 . 67 to find slope angle. maximum key blocks of a tunnel. 301 in a surface excavation. lO. 75 intersection of planes. 291 find safe slope. 49-50 sliding direction. 102 lifting. 238 comparison with vector analysis for underground excavations. 105 finding removable blocks. 66.240 Stability analysis. 169 angles between lines and planes. 302 T example of computation. 55 Symmetry. 30. 242 compared with vector analysis methods. 4I. I47 under gravity alone. 28. 203 310 united blocks. equation of a plane. 40. 150 Tunnels. 193 Stereonets. '14. 87 intersections of lines and planes. 4 slope direction. 101 find direction. 108 Small circle. underground space. l7l-74 example of calculation using vectors. 50 construction. 45-48 Stresses. 190 removability. given direction.7l v joint blocks. 99. 198. 6-8. Shi's theorem. 25 excavations. 279 Stereographic projection shape. l2l proof. 92 emptiness of a pyramid.249. 87 removability of united blocks. 147 upper versus lower focal point.l5-78 Vector analysis key blocks in a surface excavation. 53. 100. 49 Stillwater damsite. 238 removability theorem. 67-70 methods. 271 52 normal to a plane. 214 finding the JP for a given sliding direction. 4l-43. 106 half spaces.

27 WaEr radius.27 sliding) Voegele. 179-83 wall Pillars' 237 vectors Walls. M.. 296 t . 85 *"|?ltXh. W.) W surface excavations. 26 5 stereographic projection. I ndex Vector analysis (contd. two face tn\t.. 205 normal. 15 Wittke. ng (seeModes of failure.