Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. Vol. 20, No. 4, pp.

181-187, 1983 0148-9062/83 $3.00+0.00

Printed in Great Britain. All rights reserved Copyright '~ 1983 Pergamon Press Ltd

Estimation of Strength Anisotropy Using

the Point-Load Test
E. BROCH*
As most rocks are of an anisotropic nature, the measurement of the ratio
between the maximum and minimum strength, i.e. the strength anisotropy, is
of general interest. To measure the strength anisotropy by the uniaxial
compression test, core specimens drilled in different directions are needed. With
the point-load test it is possible to obtain the strength anisotropy on one core
only by first using the diametral test and then applying axial tests on the core
pieces. The paper describes results from tests performed on 33 different rocks
to find a diagram that could be used to compensate for the influence of size
and shape in the axial test. Cores were drilled both parallel and normal to
bedding or foliation planes. Analyses of the results show that the most reliable
strength index is obtained when cores are drilled normal or near normal
to weakness planes. The paper concludes with a suggested procedure for
measurements and calculations.

INTRODUCTION STRENGTH ANISOTROPY MEASURED

BY THE UNIAXIAL COMPRESSION
Rocks are in general anisotropic with regard to their TEST
physical and mechanical properties. Well defined aniso-
tropy is especially to be found among sedimentary and In the uniaxial compression test the minimum strength
metamorphic rocks, where the bedding or the foliation is obtained when the inclination of weakness planes to
normally is clearly visible. Igneous rocks may often look the direction of the major principal stress is approx. 30° .
very homogeneous and isotropic, but testing reveals that Maximum strength is obtained when the inclination is
many of their material properties vary with the direction either close to 0° or to 90 °. This means that the uniaxial
of testing. compression test gives approximately the same max-
Truly isotropic rocks are rare occurrences and should imum strength value for the two important and easily
be regarded as exceptions. In all strength testing of defined inclinations, namely, parallel to and normal to
rocks, it is therefore necessary to refer the obtained the weakness planes.
results to the relation between the direction of stresses ~Some authors find the greatest strengths in cores
and the direction of the texture of tested specimens. If drilled parallel to the weakness planes [1], while others
possible, specimens should also be tested under the same find the greatest strengths in cores drilled normal to the
magnitude and direction of stresses as the rocks will be weakness planes [2]. For six Precambrian gneisses Sven-
subjected to at the construction site. However, it is not ska V~iginstitutet [3] found that the ratios between
easy to establish future stress conditions for rock during uniaxial compressive strengths measured on cores drilled
construction. Drilling and blasting will for instance parallel and normal to the foliation were 0.92, 0.92, 0.94,
subject the rocks to very complex stress situations. 1.03, 1.08 and 1.16. This means that the variations are
Two strength values are of general interest, namely the more or less within the accuracy of the testing method.
maximum and the minimum strength of the rock. The Based on uniaxial compressive strengths measured on
maximum and minimum strengths are obtained when cores drilled parallel and normal to weakness planes, it
failure is initiated normal to, and respectively, parallel to is possible to get a false impression of an isotropic
the weakness planes of the rock, i.e. bedding, foliation, material. To obtain a correct value for the maximum
cleavage, etc. The ratio between these two extreme strength anisotropy in compression, it is necessary to test
values may be regarded as the maximum strength aniso- specimens drilled in varying directions relative to the
tropy of the rock. weakness plane of the rock. Such a procedure is both
tedious and expensive.

THE POINT-LOAD STRENGTH

ANISOTROPY INDEX
* Department of Geology,The Norwegian Institute of Technology, For the point-load test, as for other strength tests
University of Trondheim, N-7034 Trondheim, Norway. where the specimen fails in tension, the problem of
181
182 BROCH: E S T I M A T I O N OF S T R E N G T H A N I S O T R O P Y

angle between the weakness planes in the rock and the

e~
direction of the core does not exceed certain values.
In the testing of rock strengths, specimen preparation
is normally a time consuming and costly procedure. One
of the great advantages of the point-load strength test is
that it does not require machined specimens. As long as
the influence of specimen size and shape are considered

I0
\ in the calculation of the strength index, any piece of
rock, whether the surface is smooth or rough, can in
principle be tested. Using only one core sample to obtain
both the maximum and the minimum strengths of the
8 ° ~ B
rock is probably one of the fastest and cheapest methods
.-~.~ in rock engineering at present.

| RESULTS FROM AXIAL TESTS

.LI I I I I "'~l"--'--f
PERFORMED ON CORE DISCS
15 30 45 60 75 90
It has been shown [5] that test results obtained from
a (*)
axial and diametral tests on two different types of rock
Fig. 1. Diametral point-load strmagthindex, I,, as a functionof the with three different diameters were identical provided
angle, =, betweenthe foliationplane and the core axis for two gneisses that, in the axial test, specimens with a length/diameter
and a mieasehist,from Aagaard [4].
ratio of 1.1 ___0.02 were used. Based on this, the ISRM
Commission suggested that the axial test should be
defining the maximum and minimum strengths is consid-
performed on core specimens with length/diameter
erably easier. In such tests the minimum strength is
ratios of 1.1 + 0.05. This requirement is so strict that
recorded when the specimen fails along a weakness
machine cutting of specimens is in practice necessary and
plane. This has been convineingiy demonstrated by
may be one of the reasons why the axial test has been
Aagaard [4] in a dissertation at the Norwegian Institute
littled used.
of Technology. Figure 1 shows results from this work.
Read et al. [7] reported that when axial tests were
In 1972 a comprehensive description of the point-load
attempted at the suggested length/diameter ratio, the
strength test was published [5], in which a point-load
samples either rotated between the platens or broke off
strength index was defined as
at the edges. Similar features have been experienced by
P
/~ = D----~ (1) this author, but it does not always occur. For slightly
anisotropic, sound rocks it is possible to conduct valid
where P is the load required to break the specimen tested axial tests when the length of the core is greater than the
and D is the distance between the two platen contact diameter. Read et al. found after some experimentation
points. that consistent results were obtained when a ratio of 0.65
As an appendix to that paper a "Suggested Method was used. This value was adopted as standard for their
for Determining the Point-Load Strength Index" is testing.
described. This was adopted by the Commission on This empirically obtained ratio is compatible with a
Standardization of Laboratory and Field Tests of the stress analysis of cylindrical rock discs subjected to an
International Society for Rock Mechanics [6] the follow- axial, double point load by Peng [8]. His finite element
ing year. In this suggested method a strength anisotropy model shows that the stress distribution changes slightly
index Io is defined as the ratio of the corrected median for length/diameter ratios greater than 1.3 and stabilizes
strength indices for tests perpendicular and parallel to for specimens with ratios less than 1.0. He therefore
planes of weakness. In the paper the diametral and the concludes that the best specimen geometry for the axial
axial point-load test are described, and the practical use point-load is with a length/diameter ratio below 1.0.
of both the point-load strength index, Is, and the strength Similar results have been obtained by Sundae [9] in
anisotropy index, I°, are demonstrated on a core log. laboratory tests carried out on discs of three different
Since the publishing of the paper and the adoption of rocks.
the method by the ISRM, a number of authors have When performing the diametrai point-load test, the
published valuable papers on different aspects of the ISRM Commission suggests that the distance between
method and its use in practical rock engineering the contact or loading point and the nearest free end be
[7, 10--17]. It does not seem, however, that everyone is at least 0.7 d where d is the core diameter. After
aware of the potential of the method to measure max- diametral point-load testing, one is then normally left
imum and minimum strengths in one core, and thus with core pieces with lengths in the range of 0.5-1.0 d.
calculate strength anisotropies. This is done by first As the testing of rock cores in general will be per-
performing diametral tests on the core, followed by axial formed at certain standard diameters, it is desirable to
tests on the broken core pieces. As will be shown later find a best possible direct way of transforming the results
in the paper, this procedure may be used, provided the from the axial test to strength indices comparable with
 BROCH: ESTIMATION OF STRENGTH ANISOTROPY 183

I00 have been replotted in a semilogarithmic diagram as

90
80 • - Dolerite shOWn in Fig. 2. The dotted lines connnect points of the
70 same core diameter, d. It is obvious that for the range
60 of core diameters between 25 and 54ram and with
50 length/diameter ratios varying between 0.5 and 1.5, the
inclination of the lines are consistent for the two rather
40
different rocks. It has thus been possible to fit a set of
30
parallel lines like the solid lines in the diagram.
These solid lines, which establish the combined
influence of size and shape in an axial test, are replotted
20 in Fig. 3. The lines can be described by the equation
L =/(1" 10 -°°25~ (2)
where Ki is the constant defining where the single lines
IO
% are running. (K1 = I,, when D is made equal to zero). An
n 9
% example for KI = 85 is shown in the right part of Fig. 3.
8
7 Also included on these diagrams are the strength
u~ indices for the sandstone and dolerite discs for 25.5 and
6 (9
38 mm cores as a direct function of the height, D (or
length). Dotted lines parallel to the solid lines are fitted
to the plotted strength indices. The big crosses mark the
intersections between the dotted lines and the indices
X\x measured by the diametral test on cores from the same
\
P ~x\
\ x block of rock and of the same sizes, but drilled in a
direction normal to the axially tested core discs.
In these diagrams the strength indices for axially
\\ tested core discs with a diameter of 25.5 mm can be
correlated to those obtained for diametrally tested cores
when the height (or length) of the core disc is 28.7 ram.
05 ~0 15 Correlation is obtained for 38 mm cores when the height
Dld of the disc is 39.1 mm. This clearly indicates that axially
Fig. 2. Point-load strength indices, I,, obtained from axial tests on tested core pieces should have lengths slightly greater
sawn core discs of varying diameters, d, as functions of the ratio D/d. than their diameters to give commensurable strength
D is the distance between the load platens (or height of disc).
indices. This was the bails for the earlier mentioned
ISRM suggestion for the axial point load-test.
those obtained in the diametral test. Thus, when the In a recent paper, Greminger [10] reports valuable
influence of shape and size on the results from the axial results from experimental studies of the influence of rock
test ~s considered, a true strength anisotropy index can anistropy on size and shape effects in point-load testing.
be calculated. Four types of rock with varying strength anisotropy
Returning to the earlier published paper [5], the results indices (given in parentheses) were extensively tested:
from axial point-load tests on carefully sawn core discs Augen gneiss (Ia = 1.05), Ruhr-Sandstone (Ia = 1.15),

6
5 215 5 mm Cores 38ram Cores

4 I
3
"~" O~/er'~/re'-.. --. , I0 "0 Oo TM
i
"
2
- "" -.Y~o ^ . ~ - - "~ "-L.

o
n

8
7
6
-

4 - ~ X

3
- ""-, I
20 30 20 30 40
D(mm) D(mm)
Fig. 3. I, for axially loaded core discs as a function of the height, D. See text for further explanation.
184 BROCH: ESTIMATION OF STRENGTH ANISOTROPY

Chiandone-Gneiss (I~ = 2.80) and Nuttlar-Slate diagram expressed by a slightly curved line (short
(Ia = 5.70). Like other research workers [5, 9, 11], Grem- dashes) which to a large extent is subparallel to the
inger finds that the best fit with experimental data is straight equation (2) lines.
obtained by a parabolic function of strength, I,, against With the chosen a and b values, equation (6) can be
size, D. rewritten as
L ~c D-° (3) P oc ( d . D ) °75 ~ A °-75 (7)
He finds no indication of any dependence of the ex-
ponent a on anisotropy. He therefore draws the conclu- which is identical to an equation presented by Brook
sion that the size correction factor is independent of the [12, 13]. In several papers he has argued that shape and
size effect problems in the point-load test could be
degree of anisotropy and of the loading direction.
A shape effect is described by the relation overcome by introducing a strength index

P
I, oc (4) T ~ = 211.5 A0.7"---
3 (8)

Combining both size and shape effects gives where A is the loaded area. He suggests a reference area
of 500 m 2 (equivalent to diametral tests on 25 mm cores).
,,P- ~ ~ D-° (5) Objections against Brook's suggestion put forward by
Hassani et al. [14], are: (a) the area is often difficult to
Based on his experimental results Greminger chooses determine exactly and (b) the strength index as given by
a = 0.5 and b = 0.75 which gives equation (1), is so widely used and accepted that it will
be difficult to change. Recent talks with the chairmen of
/~ = K2"D -"2' (6) ISRM's Commission on Standardization of Laboratory
where K: is a constant related to the strength of the rock and Field Tests [18, 19] confirm that the Commission has
tested. no plans for changing the definition of the Point Load
To compare equation (6) with equation (2) the former Strength Index. In a revision of the "Suggested Method"
is also shown in the right part of Fig. 3 for a chosen value to be published in the near future, references will,
of K2 = 850. The equation is in the semilogarithmic however, be made to all relevant research work.

Table 1. List of tested rocks with results from diametral point-load tests performed on dry and fully water-saturated specimens. Cores are drilled
both parallel and normal to foliation or bedding
Diametral I, (MPa) with SD in ( ) Anisotropy
Spec. Geological Cores drilled parallel Cores drilled normal 1~
No. Rock type Locality era Saturated Dry Saturated Dry Sat. Dry
01 Quartz-diorite Trolla Cambr.-Silur. 7.8 (15) 12.3 (30) 4.7 (25) 5.0 (29) 1.66 2.46
02 Crystalline schist Berk/tk Cambr.-Silur. 11.6(12) 11.6(26) 5.1 (23) 6.6(31) 2.27 1.76
03 Diorite Kleft bru Cambr.-Silur. 11.5(ll) 15.5(10) 10.4(10) 14.8(9) l.ll 1.05
04 Quartzite Ringebu Eocambrian 15.5 (10) 16.0 (13) 15.4(11) 16.8 (12) 1.01 0.96
05 Arkosic sandstone Trenen Eocambrian 17.5 (14) 20.3 (40) 6.3 04) 7.7 (2 l) 2.78 2.64
06 Quartz-syenite Gjeller~Lsen Permian 7.8(15) 8.6(18) 7.6(11) 8.0(26) 1.03 1.08
07 Black shale Grodalen Cambr.-Silur. 9.6 (31) 8.0 (34) 8.8 (13) 7.8 (11 ) 1.09 1.03
08 Granite Lier~sen Permian 8.5 (24) 8.7 (14) 6.7 (6) 7.9 (13) 1.27 1.10
09 Rhomb-porphyry Toverud Permian 8.0(25) 10.3 (21) 5.2 (37) 7.6 (28) 1.54 1.36
10 Siltstone Tyrihjelmen Cambr.-Silur. 12.4(27) 12.8(39) 7.8(21) 8.9(28) 1.59 1.44
11 Limestone Ringerike Kaikv. Cambr.-Silur. 6.3(18) 6.4(18) 5.1 (38) 5.5(28) 1.24 1.16
12 Quartzite Modum Precambrian 11.9(21) 12.6(13) 10.1 (14) 9.5(18) 1.18 1.33
13 Basalt Steinsskogen Permian 17.0(29) 13.7(19) 16.8(17) 11.7(29) 1.01 1.17
14 Gneiss Hambora Precambrian 9.1(24) 10.1 (31) 4.9(18) 7.1 (18) 1.86 1.42
15 Gneiss-granite Hambora Precambrian 9.3(16) 10.8(20) 9.2(10) 9.8(15) 1.01 1.10
16 Gneiss-granite Gronningen Precambrian 12.1 (10) 14.5 (10) 8.5 (17) 10.9 (12) 1.42 1.33
17 Micashist Trlssavika Precambrian 7.6 (21) 8.5 (l 8) 3.2 (39) 5.5 (26) 2.38 1.55
18 Gabbro Myrvang Cambr.-Silur. 11.5(6) 17.9(11) 9.1 (10) 14.7(14) 1.26 1.22
19 Gabbro Heggest dam Cambr.-Silur. 9.0(ll) 13.4(9) 7.3(15) 8.7(27) 1.23 1.54
20 Micashist Odtvollen Cambr.-Silur. 11.4 (21) 9.6 (38) 2.0 (33) 4.5 (56) 5.70 2.13
21 Sandstone Sveigen Devonian 18.7(20) 17.1 (22) 17.2(25) 18.7(14) 1.09 0.91
22 Quartzite Adamselv Eocambrian 17.3 (7) 14.7 (20) 1.18
23 Biotite-gneiss* Linde(jell Prccambrian 3.0 (72) 1.6 (44) 1.88
24 Quartz-syenite Lindefjell Precambrian 5.9 5.3 I. 1l
25 Granite Lindefjell Prec,ambrian 10.6 (7) 5.7 (32) 1.86
26 Gneiss-granite Lisle~t ~brian I 1.7 (8) 8.4 (l 3) 1.39
27 Gneiss-granite Lislet Prccambrian 12.3 (8) 7.3 (14) 1.68
28 Gneiss Gjora Pr~,ambtian 5.5(21) 9.3(16) 3.1 (26) 3.8(21) 1.77 2.45
29 Gneiss Driva Prceambrian 12.4(32) 18.2(15) 2.7(29) 5.1 (21) 4.59 3.57
30 Quartz-diorite* Stzren Cambr.-Silur. 9.0 (! 1) 11.6 (11)
31 Hyperite Sohar Precambrian 10.2(5) 14.5 (6)
32 Gneiss lkstvoll, Fosen Prcc,ambrian 8.8(30) 15.7(14) 8.0(15) 10.7(10) 1.10 t.47
33 Amphibolite /~t~ord Precambrian 8.7 (14) 13.3 (6) 6.5 (11) 10.2 (15) 1.34 1.30
Average values for standard deviation: 17.5 20.2 19.9 21.3
* Excluded from average value calculation.
 BROCH: ESTIMATION OF STRENGTH ANISOTROPY 185

AXIAL TESTING OF CORE PIECES Dry " X

Sot. " •
Diametral testing of rock cores will normally produce x
0
a number of core pieces with lengths shorter than their 2O-x o o OxOXx
diameter. To investigate the possibilities of employing g ~.~x
extrapolations of axial test results along fitted lines as o o:o
x~ x
o xO~O o X
shown in Fig. 3, a total of 33 different Norwegian rocks - Io Xx ~o ~ xx x
no X o oe/ ~ o x o
have been carefully tested [15]. The ELE-point-ioad
x° OoOe o o°x
tester (from Engineering Laboratory Equipment,
.,~*l O0 X
England) was used, and a core diameter of 31.5mm o
(normally referred to as 32 mm cores) was chosen. This 5 - o ~po Xx o
is by far the most used core dimension in Norway. The
testing procedure and the calculation of results were E
performed in accordance with the suggestions outlined 3- E
tO
o

by ISRM [6]. t~

Cores were drilled from large blocks both parallel to I I I

25 30 35 40
and normal to bedding or foliation planes. Tests were
D (rnm]
performed on oven dried and fully water saturated cores
[16]. Specimen number, type of rock, locality and geo- Fig. 5. /?-values for axially loaded core pieces that give a point-load
strength index equal to diametrally loaded cores plotted against the
logical era are listed in Table 1 together with the results axial point-load strength index.
of the diametral point-load test. As the table illustrates,
a wide variety of rocks are represented with point-load not as pronounced in the diametral test as in the axial
strength indices on water saturated specimens varying test. Most of the gneisses are good examples of aniso-
from 1.6 to 18.7 MPa. Strength anisotropy indices (on tropic rocks with great scatter in strength results.
saturated cores) vary between 1.01 and 4.59. In axial tests there is greater scatter of the data from
The results from the axial tests were plotted as in cores drilled parallel to the foliation than normal to the
Fig. 4. For each type of rock, both the results from cores foliation. On the other hand, in the diametral test the
drilled parallel to and normal to bedding/foliation scatter is somewhat smaller for the cores drilled parallel
planes are plotted. The results from the diametral tests to the foliation. The difference in the scatter of the
(the median values) are also plotted on the diagram with results for the different directions is clearly less pro-
the big crosses. nounced for the diametral test than for the axial test.
A study of all the diagrams reveals that the scatter of This indicates that the most reliable results for the
the test results varies considerably from rock to rock. measurement of strength anisotropy indices on only one
Rocks of high strength anisotropy tend to show a greater core should be obtained when this core is drilled normal
scatter than rocks of low anisotropy. However, this is to the foliation.

Cores drilled parallel to foliation / bedding Cores drilled normal to foliation / bedd=ng

Dry : X Dry : X
50 ~_ ~ Sat io Sat , i o- --
40 i "

30 "~ J
~... X "" . . . ~ iI
A
o
~,x "--<.
n
2o_ o-~.._'>... ~....xx ! ~ - r

"6
:> I 0
=_
==

E
5
Q.
7

2
3
_ ~"

,,
I
20 30 40 20 30 40
D(mm) D (turn)
Oiametral /s -- S a t . ' II 5 M P o , Dry " 15.5 MPa Diamefral /s - Sat. IO..4MPa , Dry 1.4 8 M P o
Spec Na 03 Rock Diorite !- OCOII/y Kloft b f u

Fig. 4. Example of a results diagram for axial point-load testing of core p i ~ s .

RMMS 2 0 / ' ~ "

186 BROCH: ESTIMATION OF STRENGTH ANISOTROPY

25 /
/
/ / o
/
• / /
/ /
+,,_
0 / /
2.0 / /
0 /
/ • //o/ Q
i
/ • /
0
"lD /

:D 1.5 /
• //
I0 °o /

o/4~/ • •
• /o •
] ,o I I
tO 1.5 2.0 2.5 1.0 ).5 2.0 2.5

/e(ll) /$111

Fig. 6. Strengthanisotropyindices measured on one core drilled parallel to (left) and normal to (right) the foliationor bedding
and correlated with the strength anisotropy index measured by two diametral tests.

To find what core piece lengths (D) give axial strength What strength anisotropy index should be regarded
indices identical to the diametral strength indices, the as the "true" one, is difficult to decide. Anisotropy
D-values for the big crosses in the diagrams are plotted measured by diametral tests on cores drilled in two
as a function of the strength, Is in Fig. 5. The data directions is favoured by the fact that the diametrai test
include results from dry and fully saturated cores. A few is the best controlled test. Favouring anisotropy mea-
strength indices are excluded either because of too few sured by a combination of diametral and axial point-
strength tests or too wide scatter in results. The average load tests on the same core is the fact that one then really
D-value for all tests is 32.5 mm. This means that for knows that it is the same material that is tested in both
31.5 mm dia rock cores the axial strength index obtained directions. For inhomogeneous and strongly anisotropic
on a core piece which is 1 mm longer than the diameter materials this is an advantage that often may fully
(or 1.03 D) should be identical to the result from the outrange the inaccuracies in the axial test. Furthermore,
diametral test. This confirms the results presented earlier testing on one core only may considerably reduce the
from the tests carried out on sawn core discs. expense of strength testing.

CORRELATION BETWEEN POINT-LOAD APPLICATION OF THE LOAD ON

STRENGTH ANISOTROPY MEASURED ANISOTROPIC CORES
ON ONE AND ON TWO CORES
When cores of anisotropic rocks are to be tested with
When the average D-values of 32.5 mm was found, it the point-load apparatus, it is important that the load be
was drawn as a vertical line in all the semilogarithmic correctly applied. To obtain readings of the maximum
diagrams (Fig. 4) and all the equivalent axial strength and minimum strength values, the load must be applied
indices for water saturated cores were obtained. Strength so that failure is initiated normal to and parallel to
anisotropy indices then were calculated for the results weakness planes (bedding/foliation). For cores drilled at
obtained from cores drilled parallel to and normal to an oblique angle to such weakness planes, Fig. 7 demon-
foliation, respectively. In Fig. 6 these indices are plotted strates the right and wrong ways of applying the load.
against the strength anisotropy indices calculated from In the diametral test, the load should be applied along
diametral tests performed on cores drilled in both direc- the shortest axis of the elliptic weakness plane. To avoid
tions. If the diametral results are regarded as the "true" the influence of the uncontrolled shape effects, core
indices, it is of interest to note the core direction which pieces shorter than the diameter should be used.
gives the best correlation. The influence of the angle between the core axis and
Not surprisingly, the strength anisotropy index ob- the weakness planes, fl, on the diametral point-load
tained on the cores drilled normal to the bedding or
foliation of the rock gives the best correlation with the
index obtained from measurements on two cores. A few mo0~t
data points deviate considerably from the regression
line. These represent rocks with pronounced anisotropic
properties, which in terms of testing are complicated
materials with inhomogeneities and rapidly changing
properties. Gneisses and micashists are examples of such
rocks. The great scatter in strength test results often
reflects the problems of selecting representative speci- Fig. 7. Right and wrong applicationof the point.loadson cores which
mens. are drilled at an oblique angle to foliation or bedding.
 -I
BROCH: ESTIMATION OF STRENGTH ANISOTROPY 187

This is consistent with the way the diametral point load

strength is calculated where the use of the median value
instead of the mean value will have a disproportionate
effect on strongly deviating results.
8
L I I I l I Acknowledgements--Financial support for these investigations was
0 15 30 45 60 75 90 received from Norges Tekniske H6gskoles Fond. The author wishes to
. thank colleagues at the Department of Geology. The Norwegian
Institute of Technology, Trondheim, for kind help and useful advice.
Fig. 8. The variation of the point-load strength index with the angle
between the core axis and the foliation planes in a micaschist. Load
applied parallel to the foliation plane, from Aagaard [4].
Received 22 December 1982; revised 25 March 1983.

strength is shown in Fig. 8. When the load is correctly

applied, the results from diametral tests are only
influenced by the angle of the weakness planes when they REFERENCES
lie between 30 and 60 ° to the axis of the core. 1. Delkotz E. J., Brown J. W., Stemler O. A. and Ce L. Anisotropy
of a schistose gneiss. Proc. 1st I S R M Congr., Lisboa, pp. 465-470
SUGGESTED PROCEDURE FOR T H E (1966).
MEASUREMENT OF POINT-LOAD 2. Donath F. E. Experimental study of shear failure in anisotropic
STRENGTH ANISOTROPY ON rocks. Bull. Geol. Soc. Am. 72, 985-990 (1961).
3. Svenska Viginstitutet. Methods and results from testing of rocks
ONE CORE for road construction (in Swedish), Medd. No. 8, Stockholm
(1928).
(1) Cores to be tested should be drilled as 4. Aagaard B. Strength anisotropy of rocks (in Norwegian). M.Sc.
perpendicular as possible to the weakness planes dissertation, 104pp. NTH, Trondheim (1976).
(bedding/foliation) of the rock with the deviation not 5. Broch E. and Franklin J. F. The point-load strength test. Int. J.
Rock Mech. Min. Sci. 9) 669-697 0972).
exceeding 30 °. 6. ISRM. Suggested methods for determining the point-load strength
(2) Cores which are not tested as soon as they are index. ISRM Commission of Laboratory Tests, Document l, pp.
drilled, should be stored under conditions which prevent 8-12 (1973).
7. Read J. R. L., Thornton P. N. and Regan W. M. A rational
the loss of water from the rock. approach to the point-load test. 3rd Australia-New Zealand Conf.
(3) Diametral point-load tests are performed and on Geomechanics, Wellington, p. 2-35-39 (1980).
strength indices calculated in accordance with the 8. Peng S. S. Stress analysis of cyclindrical rock discs subjected to
axial double point load. Int. J. Rock Mech. Min. Sci. & Geomech.
methods suggested by ISRM. Abstr. 13, 97-101 (1976).
(4) Axial point-load tests are performed on core pieces 9. Sundae L. S. Effect of specimen volume on apparent tensile
with lengths varying between 0.5 and 1.0 times the strength of three igneous rocks. USBM RI 7846, Washington
(1974).
diameter. Each calculated point-load strength and corre- I0. Greminger M. Experimental studies of the influence of rock
sponding height (or distance between loading platens) D, anisotropy on size and shape effects in point-load testing. Int. J.
are plotted in a semilogarithmic diagram as shown in Rock Mech. Min Sci. & Geomech. Abstr. 19, 241-246 (1982).
II. Guidicini G., Neible C. M. and Cornides A. T. Analysis of
Figs 3 and 4. point-load test as a method for preliminary geotechnical
(5) A straight line with a slope defined by the equation classificationof rocks. Bull. Int.Ass. Engng Geol. 7, 37-52 (1973).
I s ---- K f l0 -°°25D, as shown in the figures, is fitted to the 12. Brook N. A method of overcoming both shape and size effects in
point-load testing. Proc. Conf. on Rock Engineering, Univ. of
points. Where this line is intersected by a vertical line Newcastle, pp. 53-70. British Geotechnical Society, London
through the D-value 3% greater than the diameter, the (1977).
corrected axial point-load strength index is read. 13. Brook N. Size correction for point-load testing. Int. J. Rock Mech.
Min. Sci. & Geomech. Abstr. 17, 231-235 (1980).
(6) The point-load strength anisotropy index, Io, is 14. Hassani F. P., S¢oble M. J. and Whittaker B. N. Application of
calculated as the ratio between the corrected axial the point-load index test to strength determination of rock and
point-load strength index and the diametral point-load proposals for a new size-correction chart. Proc. 21st U.S. Symp.
on Rock Mechanics, Rolla, Missouri, pp. 543-553 (1980).
strength index (In the case of cores drilled parallel to the 15. Broch E. The point-load test and its use in engineering geology
weakness planes Io is the inverse ratio). (in Norwegian). Rept No. 2, 148pp. Dept of Geology, NIT,
Note: As the point-load strength is strongly dependent Trondheim (1977).
16. Broch E. The influence of water on some rock properties. Proc. 3rd
on the content of water in the rock [16, 17], great care I S R M Congr., Denver, Vol. 2, pp. 33-38 (1974).
should be taken to ensure that the water content is the 17. Broch E. Changes in rock strength caused by water. Proc. 4th
same during both the diametral and the axial test. I S R M Congr., Montreux, Vol. I, pp. 71-75. Balkema, Rotterdam
(1979).
Points in the semilogarithmic diagram that deviate 18. Bieniawski Z. T. Personal communication (March, 1983).
strongly from the rest of the points may be disregarded. 19. Franklin J. F. Personal communication (March, 1983).