Suggested Methods for the Quantitative Description of Discontinuities 333

4. Savage J. F. Terrestial Photogrammetry for of little consequence as failure may occur

Geological Purposes. International Training Centre for through rotation or rolling of the small rock
Aerial Survey publication, Series B, No. 33, pp. 41-53
(1965). pieces.
5. Thompson M. M. (Ed.). Manual of (c) As in the case of orientation, the
Photogrammetry 3rd edn, American Soc. of Photogramm. importance of spacing increases when other
Falls Church, Virginia (1966). conditions for deforma tion are present, i.e. low
6. Rengers N. Terrestial photogrammetry: a
valuable tool for engineering geological purposes. Rock shear strength and a sufficient
Mech. Engng Geol. 5, 150-154 (1967). number of discontinuities or joint sets for slip to
7. Calder P. N., Bauer A. & Macdougall A. R. occur.
Stereophotography and open pit mine design. 72nd Annual
Meeting Can. Inst. Min. Metall. April 1970. Preprint (1970). (d) The spacing of individual discontinuities
8. Wickens E. H. & Barton N. R. The application of and associated sets has a strong influence on
photogrammetry to the stability of excavated rock slopes. the mass permeability and seepage
Photogram. Rec. 7(37), April, 46-54 (1971).
9. Ross-Brown D. M. & Atkinson K. B. Terrestial characteristics. In general the hydraulic
photogrammetry in open pits: 1--description and use of the conductivity of any given set will be inversely
phototheodolite in mine surveying. Trans. Inst. Min. Metall. proportional to the spacing, if individual joint
(Sect. A: MM. Industry), 81, pp. A205—A213 (1972). apertures are comparable.
10. Ross-Brown D. M. Aspects of slope design in
open pit mining. PhD. Thesis, University of London (1973).
Equipment
2. SPACING (a) Measuring tape of at least 3 m length,
calibrated in mm divisions.
(b) Compass and clinometer.
Scope
(a) The spacing of adjacent discontinuities Procedure
largely controls the size of individual blocks of (a) Whenever possible, the measuring tape
intact rock. Several closely spaced sets tend to should be held along the exposure such that the
give conditions of low mass cohesion surface trace of the discontinuity set being
whereas those that are widely spaced are measured is approximately perpendicular to the
much more likely to yield interlocking con- tape. If the tape is not perpendicular, directional
ditions. These effects depend upon the bias corrections are required to obtain the true
persistence of the individual discontinuities. spacing.
(b) In exceptional cases a close spacing (b) All distances (d) between adjacent
may change the mode of failure of a rock mass discontinuities are measured and recorded over
from translational to circular or even to flow a sampling length not less than 3 m (or the
(e.g. a "sugar cube" shear zone in quartzite). thickness of the rock unit being observed if this
With exceptionally close spacing the orientation is is less than 3 m). The sampling length should

set no. I
set no. 2

set no. 3
S, = cl, sin U2
Fig. 10. Measurement of joint spacing from observation of a rock exposure.
 preferably be greater than ten times the esti-
mated spacing. The distances (d) should be
measured to within 5% of their absolute
values.
(c) The smallest angle (a) between the
measuring tape and the observed joint set is
measured with a compass to the nearest 5'.
(d) The most common (modal) spacing is
calculated from the equation:
S = dm sin 7
where d m is the most comm on (modal) distance
334 International Society for - Rock Mechanics

Fig. 11. Histogram showing modal, minimum and maximum spacings obtained from observations of the spacing of- one
9
set. Suggested descriptions given at base of histogram.

4.
C
0 • S modal = 350 mm
measured. It is helpful to present the variation in spac-
7
upper 20-30 m. Several investigators have found
•
ing6by means of a histogram, as illustrated in Fig. 11.
a
.
a fairly reliable relationship between
frequency, i.e. number of discontinuities per
Notes 6 metre, and the longitudinal or compression (P)
0
(a) The use of a measuring tape and wave velocity 17,.
0 compass is strongly recommended, but it is (f) The spacing or frequency of
0 not
4 essential if the engineering geologist is discontinuities can also be determined from
. experienced in taking these measurements using
analysis of drill core and from borehole viewing
a visual judgement. This will depend on the degree
techniques such as borehole television cameras,
of precision required. It should be borne in photographic cameras and borehole peri
mind that discontinuities such as joints may not scopes (see 11. Drill Core for details).
be sufficiently parallel in a given set to justify
great precision. Presentation of results
(b) The average value of individual modal
60
(a) The minimum, modal and maximum
spacings (S1, S, etc.) represents the average
spacing, S (min) S, S (max) should be recorded
dimension of typical rock blocks if persistence
for each disconti nuity set. The distributions can
is assumed. Other methods of representing
conveniently be presented as histograms, one
block size from observations closeof spacing moderate
are
extremely very for wide
each set (Fig. very 11). The following
extremely
given under parameter
close close 10, Block Size. wide wide
terminology can be used
(c)
spacing In any given discontinuity set,
domains with recognizably similar spacing may Description Spacing
be separated by more massive rock containing a Extremely close spacing < 20 mm
few widely spaced discontinuities. Block Very close spacing 20-60 mm
diagrams (Fig. 2b) or histograms (Fig. 11) can Close spacing 60-200 mm
be used to indicate this type of variability. Moderate spacing 200-600 mm
(d) In general, fractures caused by blast Wide spacing 600-2000 mm
damage should be excluded from consideration Very wide spacing 2000-6000 mm
when measuring the spacing of discontinuities. Extremely wide spacing > 6000 mm
(e) In cases where rock exposures are of (b) A convenient method of presenting large
limited extent, or absent, seismic refraction numbers of spacing measurements for which
techniques can be used to estimate spacing in the statistical treatment may be required is the use
of histograms, one for each set of
discontinuities. Frequency curves for each set
can be drawn on the same diagram, giving an
immediate
 Suggested Methods for the Quantitative Description of Discontinuities 335

impression of the respective modal values and disper- those discontinuities that are unfavourably
sions. (Note: using mean in place of modal spacings orientated for stability. The degree to which
may help to eliminate difficulties with samples having discontinuities persist beneath adjacent rock
multiple, poorly-defined modes, and with samples with blocks without terminating in solid rock or
modes at very small spacings, i.e. from negative terminating against other discontinuities
exponential distributions.) determines the degree to which failure of intact
rock would be involved in eventual failure.
(c) Spacing may also be expressed as the inverse i.e.
Perhaps more likely, it determines the degree
number of discontinuities per metre. This is termed fre -
to which "down-stepping" would have to occur
quency.
between adjacent discontinuities for a failure
surface to develop. Persistence is also of the
REFERENCES greatest importance to tension crack
development behind the crest of a slope.
1. Terzaghi R. Sources of error in joint surveys.
Geotechnique, 15, 287-304 (1965). (d) In the case of tunneling, failure in the
2. Weaver R. & Call R. D. Computer estimation of first instance may be a rather local affair, and
oriented fracture set intensity. Reprint Symp. on Computers in persistence across a limited number of blocks
Mining and Exploration, Tuscon, Arizona, 17 p. (1965).
3. Da Silveria A. F., Rodrigues F. P., Grossman N. may be all that is required provided that other
F. & Mendes F. Qualitative characterization of the conditions are compatible with failure, i.e. the
geometric parameters of jointing in rock masses. Proc. 1st. existence of smooth or clay filled discontinuities
Congress of the Int. Soc. Rock Mech. Lisbon, Vol. I, pp. 225- or at least three sets. Planar discontinuities that
233 (1966).
4. Halstead P. N., Call R. D. & Rippere K. H. can be traced without offset for 5-10 m in a
Geological structural analysis for open pit slope design, tunnel construction may be of major
Kimberley pit, Ely, Nevada. Reprint, Annual AIME meeting, significance to stability, while being of minor
New York, 25 p (1968). importance in the case of a 100 m high rock
5. Ward W. H., Burland J. B. & Gallois R. W. slope or large dam abutment.
Geotechnical assessment of a site at Mundford, Norfolk, for a
large proton accelerator. Geotechnique, 18, 399-431 (1968).
6. Broadbent C. D. & Rippere K. H. Fracture studies (e) Frequently, rock exposures are small
at the Kimberley pit. Proc. Symp. on Planning Open Pit compared to the area or length of persistent
Mines, Johannesburg, 1970. Balkema, Amsterdam, 1971, pp. discontinuities, and the real persistence can
171-179 (1971). only be guessed. Less frequently it may be
7. Piteau D. R. Geological factors significant to the
stability of slopes cut in rock. Proc. Symp. on Planning possible to record the dip length and the
Open Pit Mines, Johannesburg, 1970. Balkema, strike length of exposed discontinuities and
Amsterdam, 1971, pp. 33--53 (1971). thereby estimate their persistence along a given
8. Roberston A. MacG. The interpretation of plane through the rock mass using probability
geological factors for use in slope theory. Proc. Symp. on
Planning Open Pit Mines, Johannesburg, 1970. Balkema, theory. However, the difficulties and
Amsterdam, 1971, pp. 55-71 (1971). uncertainties involved in the field
9. Grainger P., McCann D. M. & Gallois R. W. measurements will be considerable for most rock
The application of the seismic refraction technique to the exposures encountered.
study of fracturing of the Middle Chalk at Mundford,
Norfolk. Geotechnique, 23, 219-232 (1973).
10. Piteau D. R. Characterizing and extrapolating Equipment
rock properties in engineering practice. Rock Mechanics
(Springer-Verlag), Suppl. 2, pp. 5.31 (1973).
(a) Measuring tape of at least 10 m length.
I1. Priest S. D. & Hudson J. A. Discontinuity spacings in rock. Int.
J. Rock Mech. Min. Sci. & Geomech. Absir. 13, 135-148 (1976). Procedure

(a) Indi vi dual rock exposures, or

3. PERSISTENCE recogni sed domains, should first be
described according to the relative persistence
Scope of the different discontinuity sets present. The
sets of discontinuities can be distinguished by
(a) Persistence implies the areal extent or the terms persistent, sub persistent and non
- -

size of a discontinuity within a plane. It can be persistent respectively. Simple labelled field
crudely quantified by observing the discontinuity sketches such as those illustrated in Fig. 12, can
trace lengths on the surface of exposures. It is be useful aids in subsequent interpretation.
one of the most important rock mass (b) Efforts should then be made to measure
parameters, but one of the most difficult to the discontinuity lengths in the direction of dip
quantify in anything but crude terms. and in the direction of strike. This may be
(b) The discontinuities of one particular set impossible in the case of limited planar
will often be more continuous than those of the exposures. However, in the case of large three-
other sets. The minor sets will therefore tend to dimensional exposures such as curved open pits
terminate against the primary features, or they with benches, or underground openings with
may terminate in solid rock. intersecting tunnels, it may be possible to obtain
(c) In the case of rock slopes and dam useful size-frequency histograms for each of the
foundations it is of the greatest importance to discontinuity sets.
attempt to assess the degree of persistence of
 The modal trace lengths measured for each set can
be described according to the following scheme:
Very low persistence .< 1 m
Low persistence 1-3 m
Medium persistence 3-10 m
High persistence 1G--20 m
Very high persistence > 20 m
(c) A useful procedure during the mapping
of discontinuity lengths is to record the type of
termination according to the following
scheme. Discontinuities
336 International Society for Rock Mechanics

age length of 2.9 m had one end exposed, and 912 (24%)
with an average length of 6.3 m had no ends exposed.
(b) Analyses of dip lengths and strike
lengths performed by Robertson [4] have
indicated that discontinuities tend to be of
(• approximately isotropic dimensions. When
terminating in solid rock they may therefore
tend to be circular, and presumably rectilinear
when terminating against other discontinuities.
L / (c) Statistical tests simulating circular outline
/ discontinuities with a normal distribution of
- -7- - -7- diameters randomly spaced in the rock mass,
indicate that the mean trace length can range
from slightly smaller to slightly larger than the
mean diameter [5]. This is the result of the
greater probability of intersecting the larger dis-
continuities outweighing the fact that trace lengths
(i.e. chords) are inherently shorter than
diameters.
(d) Statistical methods can be used to
analyse the maximum lengths of discontinuities.
Using such techniques it is possible to estimate the
(• (f)
expected recurrence interval for discontinuities
of any specified length. Alternatively it is
possible to estimate the mean probability of a
discontinuity exceeding a specified length
occurring in any portion of the rock mass.
For example, if after analysis it is found that
major-discontinuities with strike lengths of 50 m
tem
or more are spaced on the average at 150 m, it is
possible to estimate the probability of strike
Fig. 12. Simple sketches and block diagrams help to indicate the
relative persistence of the various sets of discontinuities. Examples
lengths of 50 m or more occurring in any 100 m
adapted from [1] and [2]. interval measured normal to the strike. The
probability is equal to IN = 0.66. If the complete
which extend outside the exposure (x), should be differ- distribution of sizes is known (Procedure (b)), the
entiated from those that visibly terminate in rock in probability of occurrence of a discontinuity of a
the exposure (r), and from those that terminate against certain size can be evaluated on the basis of
other discontinuities in the exposure (d). A systematic extreme value statistics. A useful example of its
set, of discontinuities with a high score in (x) is application to rock slope stability analysis is
obviously more persistent than a sub-systematic set with given by McMahon [6]. Note that the ill-
predominant scores in (d). Non-systematic discontinui- defined lower bound to observations of trace
ties will tend to have highest scores in (r). length (inevitable if the shortest features are
(d) Termination data (x, r or d) should be recorded ignored) leads to underestimation of the
for each end of the relevant discontinuities, together frequency of discontinuities and overestimation
with the length in metres. (Example: 8(dx) = discon - of their size.
tinuity length of 8 m, one termination against another (e) The descriptive term persistence may in
discontinuity other termination invisible because feature theory be quantified by defining it as the
extends beyond the limits of the exposure). It is impor- percentage of the total area of a plane through
tant to specify the dimensions of the exposure on which the rock mass which is formed by
measurements were made since this will obviously in- discontinuities coincident (co-planar) with this
fluence both the number of (x) observations and the reference plane. In practice, waviness of most dis-
relevant lengths. continuities frustrates strict interpretation. A
practical alternative is to select a band width
Notes equal to the mean spacing of the discontinuities
(a) Piteau [3] has demonstrated that discontinuities in the particular set, and to estimate the
where both terminations can be seen are generally persistence within this reference band. Since, on
smaller than discontinuities where one or no termina - a probability basis, only one discontinuity
tions can be seen. In a sample of 3844 joints at the would be expected to occur within this band, a
Nchanga Mine, 1394 (36%) with an average length of slightly more realistic estimate of persistence is
1.4 m had both ends exposed, 1538 (40%) with an aver- obtained.
 (1) When assessing the persistence of the various dis-
continuity sets it is important to investigate the possibi
lity of a stepped failure surface forming, as illustrated
by failure modes (2). and (3) in Fig. 13. This mode of
failure may tend to occur when the set involved in
shear has less than 100% persistence. Downstepping
will tend to develop such that only a minimum percent -
age of the resulting shear surface passes through intact
rock. The persistence of a potential failure surface will
normally be higher than that along planes or bands
 Suggested Methods for the Quantitative Description of Discontinuities 337

0
intact

Syr'dyes

‘31;
0 pl a n e f a i l u r e ® stepped failure stepped 5'

failure
`21:i

intact bridges
Fig. 13. Idealized examples of potential failure planes showing the importance of "intact bridges" and "down-
stepping".

su b - p e r si st e n t se t

non-persistent
set
persistent set

Examples adapted from [4] and [7].

parallel to a single set, unless the latter have in doubt, since the above cohesion is
100% persistence. usually one to two orders of magnitude
(q) Estimates of persistence for given greater than the shear strength of the
planes, bands or specific failure surfaces discontinuities.
have at present to be based on engineering Presentation of results
judgement and should be purposely
(a) The various sets of
weighted in the direction of conservatism
discontinuities should be described as
(i.e. closer to 100% persistence since the
systematic, sub-systematic or non-
shear strength of the intact rock bridges will
systematic according to their relative
form a dangerously high percentage of the
persistence. Block diagrams or
total shear strength of the compound failure
photographs should be labelled
surface. The shear strength (cohesion) due to
accordingly.
any intact rock bridges can be crudely
(b) Where exposures are of
estimated from the following relationship
suitable dimensions, size-frequency
which is derived from the Mohr diagram,
histograms of trace lengths observed
assuming a linear shear strength envelope:
for each set of discontinuities should
c = at)1 be given. (This is necessary if
where: probability theory is to be applied
ac = uniaxial compressive strength of the subsequently). Mean trace lengths
intact rock a t = tensile strength of the (in both strike and dip directions)
intact rock. should be quoted.
If it is assumed for simplicity that (7./cr, = 9, (c) Termination data which has
then the cohesive strength is equal to one been recorded for each discontinuity
sixth of the unconfined compressive strength. sampled (e.g. 8dx), should be pres-
It is safer to assume 100% persistence when ented in the form of a termination
 index (T r ) for the rock mass as a
whole, or for chosen domains. T. is
defined as the percentage of the
discontinuity ends terminating in
rock (Er) compared to the total
number of terminations (Er + Ed +
Ex). The latter is equal to twice the
total sample since each trace has
two ends.
(Er) x 100
T=
2(no. of discontinuities observed
(It is to be hoped that systematic
collection of data
338 International Society for Rock Mechanics
concerning Ty through application of these ISRM Sug- 11. Nickelsen R. P. & Hough V. N. D. Jointing in the
Appalachian Plateau of Pennsylvania. Bull Geol. Soc. Am. 78,
gested Methods will eventually improve the estimation 609-63O (1967).
of persistence).
12. Halstead P. N., Call R. D. & Rippere K. H.
(d) The persistence of potential failure surfaces (in- Geological structural analysis for open pit slope design._
cluding stepped surfaces) should be estimated, if this Kimbley Pit, Ely, Nevada. AIME, Preprint No. 68-Am 85
is appropriate to the project being investigated. The (1968).
13. Piteau D. R. Geological factors significant to the
estimate should perhaps be rounded upwards, to the stability of slopes cut in rock. Symp. on Planning Open Pit.
next multiple of 10% (i.e. 92% is assumed to be 100%). Mines, Johannes burg, 1970. Balkeina, Amsterdam 1971,
pp. 33-53 (1971).
14. Babcock E A. Regional jointing in Southern
REFERENCES Alberta. Can J. Earth Sci. 10, 1769-1781 (1973).
15. Bernaix J. Properties of rock and rock masses..
1. Muller L. Der Felsbau. Ferdinand-Enke-Verlag, (General report.) Proc. 3rd. Cow. Int. Soc. Rock Mech. Denver.
Stuttgart, 624 p. (1963). Advances in Rock Mechanics, Vol. IA. pp. 9-38 (1973).
2. Price N. J. Fault and Joint Development in Brittle
. 16. Cruden D. M. Describing the size of
and Semibrittle Rock. Pergamon, Oxford, 176 p. (1966). discontinuities. Int. J. Rock Mech. Min. Sci. & Geomech.
3. Piteau D. It. Characterizing and extrapolating Abstr. (Pergamon) 14, 133-137 (1977).
rock joint properties in engineering practice. Rock
Mechanics (Springer-Verlag), Suppl. 2, pp. 5-31 (1973). 4. ROUGHNESS
4. Robertson A. MacG. The interpretation of
geological factors for use in slope theory. Symp. on Planning Scope
Open Pit Mines, Johannesburg, 1970. Balkema, Amsterdam
1971, pp. 55-71 (1971). (a) The wall roughness of a discontinuity is
5. Barton C. M. An analysis of rock structure and a potentially important component of-its shear
fabric in the CSA Mine, Cobar, NSW. CSIRO Division of
Applied Geomechanics, Tech. Paper No. 24 (1976) strength, especially in the case of undisplaceci
6. MtMahon B. K. Design of rock slopes against and interlocked fea tures (es. unfilled joints).
sliding on preexisting fractures. Proc. of 3rd. Conga of Int. The importance of wall roughness declines as
Soc. Rock Mech. Denver, Advances in Rock Mechanics, Vol.
IIB, pp. 803-808 (1974). aperture, or filling thickness, or the degree of
7. Jennings J. E. A mathematical theory for the any previous displacement increases.
calculation of the stability of slopes in open cast mines.
Symp. on Planning Open Pit Mines, Johannesburg, 1970,
(b) In general terms the roughness of
Balkema, Amsterdam 1971, pp. 87-102 (1971). discontinuity walls can be characterized by a
8. Hodgson R. A. Regional study of jointing in the waviness (large scale undulations which, if
Comb. Ridge-Navajo Mountain area, Arizona and Utah, Bull.
Am. Ass. Petrol Geol. 45, 1-38 (1961). interlocked and in contact, cause dilation during
9. John K. W. An approach to rock mechanics, J. shear displacement since they are too large to
Soil Mech. Fdns Div. Am. Soc. civ. Engrs SM4, 1-30 (1962). be sheared off) and by an unevenness (small
10. McMahon B. K. Indices related to the
mechanical properties of jointed rock. Proc. of 9th. Symp. scale roughness that tends to be damaged during
on Rock Mech., Status of Practical Rock Mech., Ch. 6, pp. shear displacement unless the discontinuity walls
117-133 (1967). are of high strength and/or the stress levels are

1 La bo ra to ry
sheer toot
2 In situ
'boar toot

Fig. 14. Different scales of discontinuity roughness are sampled by different scales of tests, Waviness can be characterised
by the angle (i).
low, so that dilation can also occur on these
small scale features).
(c) In practice waviness affects the initial
direction of shear displacement relative to the
mean discontinuity plane, while unevenness
affects the shear strength that would normally be
sampled in a laboratory or medium scale in situ
direct shear test (see Fig. 14).