Journal of the Geological Society of Sri Lanka, Vol.

14, 65-69

Journal of the Geological Society of Sri Lanka

Vol. 14 (2011): 65-69.
C.B.Dissanayake Felicitation Volume

Correlations between Some Strength Properties of

Metamorphic Rocks of Sri Lanka

U.de.S. Jayawardena
Department of Civil Engineering, University of Peradeniya, Peradeniya, Sri Lanka.
(email: [email protected])

ABSTRACT

In order to find an inexpensive and quick site investigation method for determination
of strength of in-situ rock masses, a laboratory study was carried-out on some strength
properties of fresh rocks of Sri Lanka. Using borehole and irregular fresh rock samples,
Point Load Strength, Tensile Strength and Compressive Strength were determined for
common Sri Lankan metamorphic rocks. The relationship found between these three
properties indicates that the approximate knowledge of one strength property is sufficient
to find the other two for different rocks. This is one of the most economical and fast me-
thods useful during preliminary site investigations.

INTRODUCTION their mechanical properties. A classification of

the various strength properties of Sri Lankan
One factor necessary for designing and eva- rocks is therefore an urgent necessity because it
luating the stability of any engineering structure gives prior knowledge of these properties by
is knowledge of the mechanical properties of its identifying the type of rocks in the field.
material or knowledge of how the material be- There are many strength properties. Some
haves under the action of applied forces. The properties of rocks can be measured by direct
composition of most structural materials such methods. There are indirect methods to meas-
as metals and concrete is uniform and repro- ure the same property in different ways. The
ducible. However, the composition of even the indirect methods are quick, economical and
more common rock types is highly variable. This need less man power. The indirect results are
may be due to various factors such as rock ani- related to the direct results both in field tests
sotropy, cementing materials, mineralogy, rock and laboratory tests. Development of any rela-
structure and texture, geological actions, altera- tionship between laboratory and field tests will
tion and state of decomposition of a given rock be useful for future investigations.
(Goodman, 1976). Therefore, different rock The main objective of the present study is
types give rather different values for the same to make a useful contribution to the literature
property. This indicates that the determination on the strength properties of Sri Lankanrocks.
of the strength properties of rocks in relation to The author carried-out a programme of re-
geotechnical problems is an essential part of search to determine these properties and some
most geotechnical investigations (Bell, 1980). of the results have been already published
The various ways of determining these proper- (Jayawardena, 2001). This paper highlights the
ties of rocks practiced by engineering geologists values of different strength properties and in-
are: by referring to values given in the appropri- ter-relationships among those and the devel-
ate literature, by in-situ experiments and by a opment of empirical relations between strength
programme of laboratory testing. properties of Sri Lankan rocks.
As far as Sri Lankan rocks are concerned,
there is very limited data available for use as
literature reviews especially with reference to

65
Jayawardena, Some Strength Properties of Metamorphic Rocks

Summary of the Geology of Sri Lanka discs using a laboratory rock cutting machine in
Geologically nine tenths of Sri Lanka is such a way that the thickness of each specimen
made up of high-grade metamorphic rocks of is equal to the half of the diameter (27 mm).
Precambrian age i.e., older than 570 million Brazilian test described by ISRM Committee on
years, belonging to one of the ancient and sta- laboratory tests (1977) was performed for each
ble parts of the earth's crust, called the South specimen using an apparatus specially made for
Indian Shield. The remaining rocks are sedimen- this (by way of steel jaws). The minimum num-
tary rocks of predominantly Miocene age in the ber of samples per test was five.
north west (and very few places of the south- Uniaxial Compressive Strength test: This
east) with some Jurassic sediment preserved in preparation was in accordance to ASTM (1977,
small faulted basins. There are recent sedimen- 1972). Samples were carefully checked to select
tary formations, identified as Pleistocene Depo- the most suitable specimens for testing. Sam-
sits in a few locations. Some granites, dolerites, ples with joints, cracks, and cavities and with
pegmatites, quartz veins and a carbonatite are little irregularities were rejected. The samples
the intrusive rocks in Sri Lanka (Cooray, 1984). were tested using a 100 ton Universal Testing
Charnockitic gneiss or charnockite, quartzite, Machine, AVERY, England No 62351 with a
marble, dolomite, granulite, migmatite, gneisses spherical seating top cap to ensure proper con-
(garnet sillimanite graphite gneiss, hornblende tact. Only one sample was used per test. Table 1
biotite gneiss, biotite gneiss, calc gneiss, cordie- shows the number of samples used for each
rite gneiss, wollastonite-scapolite gneiss, granit- test.
ic gneiss) and amphibolites are the common
Precambrian metamorphic rocks in Sri Lanka. RESULTS
These rocks have been formed under the granu-
lite-facies or amphibolites-facies of regional me- The major rock types selected for this study
tamorphism and deformation. were charnockitic gneiss or charnockite, garnet
sillimanite gneiss, hornblende biotite gneiss,
biotite gneiss, migmatite, quartzite, granulite
EXPERIMENTAL METHOD
and marble which are the major metamorphic
rocks in Sri Lanka. Some other rock types,
Sample Collection, Preparation and Laboratory
namely, amphibolite, calc gneiss and pink
Experiments
feldspar gneiss were also selected though those
rocks are not widely distributed in the country.
Borehole samples were obtained from vari-
The summary of results of each property de-
ous drilling sites belonging to different engi-
termined is given in Table 2.
neering projects. All samples were NX size (ap-
proximately 54 mm) (AIT, 1981). All collected
Brazilian Tensile Strength (BTS): The strongest
samples were fresh and represented different
rock, which can hold a much higher tensile
rock types. No weathered samples were se-
force, is charnockitic gneiss and the weakest
lected for this experiment. Irregular fresh rock
rock is the marble.
samples were collected from different quarries
and outcrops of selected localities. All samples
Uniaxial Compressive Strength (UCS): The high-
were collected from the major rock types in the
est value of 231.6 MPa was obtained from
country and the following tests were carried out
charnockitic gneiss and the lowest was from
in the laboratory.
marble rock. In addition to that, charnockitic
Point Load Strength index test: Rectangular
gneiss has wider range of uniaxial compressive
block specimens were prepared in various sizes
strengths than the other rocks.
for the tests, according to the recommenda-
tions made by the ISRM (1985). The method
Point Load Strength (PLS): Point load strength
suggested by ISRM (1985) was applied for tests
test gives an index and provides a measure of
and calculations. Axial tests were carried out for
the quality of the rock (from which the strength
borehole samples. About 20 samples were pre-
is estimated). But it does not measure the
pared per test.
strength of the rock. About 20 samples of dif-
Brazilian Tensile Strength test: Cylindrical
ferent size from one rock type at a selected lo-
samples of hard rocks were cut into circular
cation were used to draw the graph between

66
 Journal of the Geological Society of Sri Lanka, Vol. 14, 65-69

Table 1 Number of tests carried out for different rock types (b=borehole samples, r=block samples).

Rock type Point Load Strength Brazilian Tensile Uniaxial Compres-

Index test Strength test sive Strength test
Charnockite/ 138 (95b+43 r) 95 95
Charnockitic gneiss-CHA
Garnet sillimanite graphite 17 (11b+6 r) 11 11
gneiss-GSG
Hornblende biotite gneiss- 42 (19b + 23 r) 19 19
HBB
Biotite gneiss-BTG 58 (44b + 14 r) 44 44
Quartzite-QTZ 26 (12 b + 14 r) 12 12
Marble-MBL 37 (21 b + 16 r) 21 21
Granulite-GRN 7 (5 b + 2 r) 5 5
Migmatite-MIG 5 * *
Microcline gneiss-PNG 4 * *
Calc gneiss-CLG 1 * *
Amphibolite-AMP 1 * *

the failure load and the effective sample diame- Statistical analysis
ter. All are either rectangular block samples or A simple statistical method (Gupta and Kapoor,
borehole samples and hence the area calcula- 2002) was used to find out the variations be-
tion was very accurate. Then the failure load for tween the ratios of Uniaxial Compressive
a 50 mm diameter sample was obtained from Strength and Point Load Strength index, and
the graph to calculate the point load strength between Brazilian Tensile Strength and Point
index. Similar to Uniaxial Compressive Strengths Load Strength index for different rock types. For
and Brazilian Tensile Strengths, charnockitic this comparative study only the results obtained
gneiss shows the highest Point Load Strength from the borehole samples were used. Samples
index and marble has the lowest. Point Load from one rock type within one core sample box
Strength index was measured perpendicular to (same borehole, within one meter length at se-
the mineral bands of all banded rocks. Generally lected depth) were used for these three tests
the index parallel to the mineral bands in those and the same results were used to calculate the
rocks is slightly less than the values measured in ratios, mean values and standard deviations.
perpendicular directions. The highest index val- There are no relationships among the sampling
ue was used for comparison with the other depth, ground mean sea level or depth of sam-
rocks. pling point below the bedrock level. Point Load

Table 2: Ranges of Point Load Strength index, Brazilian Tensile Strength and Uniaxial Compressive Strength of
different rock types in Sri Lanka.

Rock type Point load strength Brazilian tensile Uniaxial compressive

index, MPa strength MPa strength MPa
Charnockite/ 6.8-16.8 7.6-15 104.2-231.6
Charnockitic gneiss-CHA
Garnet sillimanite graphite 4.4-7.5 5.2-9.2 75.1-116.3
gneiss-GSG
Hornblende biotite gneiss-HBB 3.5-7.6 4.1-8.8 60.3-125.5
Biotite gneiss-BTG 4.4-10 5.2-10.8 74.9-159.2
Quartzite-QTZ 4.8-7.6 5.2-8.3 70.2-118.4
Marble-MBL 2.0-6.9 2.3-8.4 33.8-124.3
Granulite-GRN 4.5-5.2 5.5-6.2 81.4-90.7
Migmatite-MIG 5.0-7.8 * *
Microcline gneiss-PNG 10.8-12.2 * *
Calc gneiss-CLG 7.4 ** * *
Amphibolite-AMP 11.2 ** * *
*No borehole samples from these rock types; ** Tested only one sample. This should not be used for further discussion.

67
Jayawardena, Some Strength Properties of Metamorphic Rocks

Table 3: Relationship between Uniaxial Compressive Strength (UCS), Brazilian Tensile Strength (BTS) and Point
Load Strength index (PLS) of different rock types in Sri Lanka (mean value and standard deviation).
Rock type UCS UCS BTS
BTS PLS PLS
Charnockite/Charnockitic gneiss 14.73 + 1.19 17.14 + 1.55 1.16 + 0.04
Garnet sillimanite graphite gneiss 14.02 + 1.02 16.73 + 0.88 1.20 + 0.57
Hornblende biotite gneiss 14.79 + 1.14 17.00 + 0.88 1.16 + 0.11
Biotite gneiss 14.45 + 0.90 17.01 + 1.20 1.18 + 0.33
Quartzite 14.43 + 0.66 15.70 + 0.42 1.09 + 0.04
Marble 14.52 + 1.19 17.77 + 0.55 1.23 + 0.09
Approximate average 14.49 + 1.01 16.89 + 0.89 1.17 + 0.19

Strength index found from the rectangular block approximate average for all rocks the ratio be-
samples were not considered because those are tween Uniaxial Compressive Strength and Point
from different localities and no Uniaxial Com- Load Strength index given in Table 3 for differ-
pressive Strength and Brazilian Tensile Strength ent rock types can be applied. In general, the
for the rocks are available from a given location. Uniaxial Compressive Strength can be consi-
Table 3 gives the mean values and standard dered as 17 times of Point Load Strength index
deviations of these ratios for major rock types. for any rock type.
The project sites from where the rock sam-
DISCUSSION ples were collected for the experiments are lo-
cated in different parts of Sri Lanka and, are not
The experiments carried out by Lumb very close to each other. And also the rock
(1982) on Hong Kong granite showed that the types selected for the experiments are the ma-
Uniaxial Compressive Strength was 22 times the jor rocks in the country. Therefore, the ratios
point load strength index and 14 times of Un- found by this investigation can be used easily
iaxial Tensile Strength on average. Vutukuri & for any engineering project in the future.
Katsuyama (1994) showed that on average the The Point Load Strength index test, which is
Uniaxial Compressive Strength is 20-25 times considered as a field test can be carried out on
Point Load Strength index. However, the ratio irregular shaped specimens. The equipment is
can vary between 15 to 50 times for many dif- portable and can be moved out to the field very
ferent rock types. According to Goodman easily. Hence, laboratory facilities are not ne-
(1976), the Uniaxial Compressive Strength is cessary. Therefore, the preliminary investiga-
approximately equal to 24 times the Point Load tions can be done in the field itself. The field
Strength index. investigations of rock strength by Point Load
Table 3 shows the results of statistical anal- Strength index give an approximate evaluation
ysis among the different properties. The ratios and classification of the intact rock. This indirect
do not represent the banded rocks or non- test is quick, economical and one person can
banded rocks. The vertical drilling through rocks handle it. It can be noted that, this is a very
was not either parallel or perpendicular to the good field classification and it is a simple me-
foliation plane. Therefore, the loading direc- thod to correlate with other properties. The
tions to the specimens were not either parallel ratios of uniaxial compressive strength, Brazilian
or perpendicular to the mineral bands. Howev- Tensile strength with Point Load Strength index
er, the ratio of Uniaxial Compressive Strength to found by this investigation for the Sri Lankan
Brazilian Tensile Strength does not show very rocks will be very useful in the future engineer-
wide range and it can be considered as uniform. ing projects of the country.
Therefore the Uniaxial Compressive Strength
generally for all rocks is 14.5 times of Uniaxial CONCLUSION
Tensile Strength. The ratio between Uniaxial
Compressive Strength and Point Load Strength The general range of the values of some
index occurs between 15.7 and 17.77. This mechanical properties was found for major me-
range is little wider than the ratio between Un- tamorphic rocks (fresh) in Sri Lanka. These re-
iaxial Compressive Strength and Brazilian Ten- sults will be very important to get a prior know-
sile Strength. Therefore rather than using an ledge about the strength properties of rocks

68
 Journal of the Geological Society of Sri Lanka, Vol. 14, 65-69

before the investigation stage. Therefore, these Bell, F.G. (1980) Engineering Geology and Geo-
results will be very useful contribution to the technics, Newnes-Butterworths, London, 497
literature on the strength properties of Sri Lan- pp.
kan rocks and this paper can be considered as a Cooray, P.G. (1984) An Introduction to the Ge-
useful reference for the engineers in Sri Lanka. ology of Ceylon, 2nd edition, Colombo, Dept. of
The Uniaxial Compressive Strength of the National Museums, Govt. Press, 340 pp.
metamorphic rocks in Sri Lanka is equal to 17 Goodman, R.E. (1976) Methods of Geological
times of Point Load Strength and 14.5 of Brazili- Engineering in Discontinuous Rocks, West
an Tensile Strength. Hence the measurement of Publishing Co., New York.
Point Load Strength index is enough to get the Gupta S.C. and Kapoor V.K. (2002) Fundamen-
approximate values of Uniaxial Compressive tals of mathematical statistics, New Delhi, Sul-
Strength and Brazilian Tensile Strength of the tan Chand & Sons.
fresh parent rock. This is the simplest low-cost International Society for Rock Mechanics (ISRM)
method to find the strength of any rock within a (1977) Suggested Method for Determining
short period. This rough estimation may be very Tensile Strength of Rock materials, ISRM
helpful for the preparation of preliminary esti- Committee on Laboratory and Field Tests, Int.
mating and designing purposes for any geo- Jour. Rock Mech. Min. Sci. & Geomech. Abstr.
technical engineering project. Further studies in 15(3): 99-103.
similar direction should be undertaken with International Society for Rock Mechanics (ISRM)
more data, so that better correction can be ar- (1985) Suggested Method for Determining
rived. Point Load Strength, ISRM Committee on La-
boratory and Field Tests, Int. Jour. of Rock
REFERENCES Mech, Min. Sci. & Geomech., 22 (2): 53-60.
Jayawardena, U.de S. (2001) A Study on the en-
American Society for Testing of Materials gineering properties of Sri Lankan Rocks, Jour.
(ASTM) (1977) Standard Test Method for Di- of the Inst. of Eng., Sri Lanka xxxiv: p 7-21.
rect Tensile Strength of Intact Rock Core Spe- Lumb, P. (1982) Engineering properties of fresh
cimens, ASTM Designation, D 2936-71: ASTM and decomposed igneous rocks from Hong
Standards Part 19, 381-388. Kong. Eng. Geol., 19: 81-94.
American Society for Testing of Materials Vutukuri V.S, and Katsuyama, K. (1994) Intro-
(ASTM)(1977) Standard Test method for Un- duction to Rock Mechanics. Tokyo, Industrial
confined Compressive strength and Intact Publishing & Consulting, Inc.
Rock Core Specimen, ASTM Designation, D
2938-71a, ASTM Standards Part 19: 389-390.
Asian Institute of Technology (1981) Laboratory
Manual for Rock Testing, Bangkok, Thailand,
205 pp.

69