Professional Documents
Culture Documents
SR Hencher & LR Richards (1989)
SR Hencher & LR Richards (1989)
proced~~.
es is given b " ie htternatio nai shear strength values
v of sou,
ety for Rock M
e advice on mte
g des provide
fofiowing ~
Thh PaPer describe and discusse s e
.
ete sta gesofdirectsh
of whichis unpoItant for Mi'tmg,each
Practice are
arly withr
me~emmentofr
l
~~tothe
i
tafion and m
wed rigorousi y
ve tical and
~ ~
d accurate mtterpretation of re ul ts.
phng
~entation
"o 'ntal strain ~ Testing
ween iaborat
~ ~y is
~ Lterpretation
shear stre 'c field values
is much less st Fundament~ f shear
stren@h
y Performed or
~ ~ect shear
s there are si leant differ
doctm,e ted or imposslb
Laboratory to mterpret; cor gently many
testm geotechni~ gineers have httl
confide "cein rock ah ar st data and
%afro'iscontinuities
"
rely mstead on e"Perience, ' item~ e
Plane of we~
g neral term for
ess inrock f
review and or empir.~
byy SR Henchert & LR RichardsO . ds such ex Ple, a jomt bed . g Plane,
of BaIton (igZ s
rise to I
d'proach
erableel 0 or sch
ol'leavage
~em~
tiFeis--
s
ilyteee
i
tete) ~ I iirlrr
- ss - ~S Iilai'Nr~ ~
~ s. ice -MMasstu~
s ~ eesM
i see
- See S
'
i.l
e:~ i ~
BASIC FRICTION
~
Introduction NON Dl LATING SURFACE
~ IPO DEPENDS ON
TEXTURE AND MINE
The asse
assessment of the RANGING FR
FROM Ir'OR
kdis o
's IFICIALLY SMOOTH EO
SURFACES TO MOe
XYS'ifC, NO,5
IP H4y, dcylh o-yH8o~~, uiDo.S~Aoo
2.IKXIGHNESS
A == »»'-
I
-- --~~ J 'A'
1~~ (oculo-~ A
B
C
B'~rgreogo
ot
sliding
1»
-iryrsm,
IdoryffI Aofiscg
rts.f
IPsNUIT
'&aNIÃ6, oAsNrf+
coioNe.~ lseso
of oPsR Eorm s«ILIUII k»
~~ ydlfI siysNri.
fzrs ruL.
a5P
Sgdj~
~HrwroilI
IIM'essaK
3.SLSIRrrCE
ANNOTATED
DESCRIPTION
DIAGRAMS PHOTOGRAPHS
gtrRrt
Syfikkk,
. Q'cssIHlsrIRS BN
I Issr
«iud'o . Afirorgst
Uirr
~, Ltsotroj eI
IeINVIMR Uorsr'fs
dsfA reek.
VhVNIVNUA «UHA
3. OsccNTNUITY DEscRPTKss
~Q M d4. Itu..
4. SAsPLE aMENSlae
BOTTOM
~ ~ BOTTOM
BIRECT SHEAR T EST S SAMPLE NUNER BIRECT SHEAR TE STS SAMPLE NUMBER
Fig.5. Example of sample documentation sheet. Fig.6. Example of sample illustration sheet.
Wherever possible, mating joint surfaces testing needs to he recorded together roughness of the surfaces can be
should be selected for testing if with all pertinent details of the test recorded using a simple proGlograph as
representative of the in situ discontinuity programme itself. Typical proformae for illustrated in Fig.7. These measurements
although non mating surfaces will often sample descriptions are given in Figs. 5 are taken for documentation purposes
yield similar basic friction values after the and 6. The descriptions need not be over only as the effective roughness must be
corrections for roughness. Even allowing detailed but should he sufGcient to record calculated from dilation measurements
for dilational effects, mating surfaces will those factors that will influence the test during the test.
sometimes give higher corrected results. The most important descriptions
strengths. This is generally due to close are those relating to surface mineralogy Test procedure
textural interlocking (non dilational) in and minor morphology.
matched surfaces. Following initial sample set up and
After testing, the surfaces should be descriptions, the shear box should be
Samples must be cut to size to fit the shear examined, particularly with regard to the assembled and the first normal load
box and set in dental plaster or some other conditions of the areas involved in applied. The normal stress range should
suitable fixing medium. Kaffir D dental shearing. It is good practice to be selected to simulate the Geld
plaster available from British Gypsum is photograph surfaces before and after conditions. If tests on wet surfaces are
used at Leeds University. The optimum shearing using low angle lighting to required, a dam of modelling clay can be
mix for this plaster is 100 parts plaster to 35 emphasise the relief. The general built around the lower hox and water
parts water (approximately 1.5kg plaster introduced using a squeeze bottle.
to 0.51itres water). The plaster sets rapidly
and achieves a compressive strength of Testing can be carried out by either single
23MPa after 2 hours and 40MPa after 12 or multistage procedures. For single stage
hours. The procedure for setting up tests, each sample is sheared at a constant
samples is illustrated in Fig.4. normal load. The aim is to obtain both
peak strength and residual strength
Documentation values. The residual strength is not a
material constant but is a function of the
One of the weakest aspects of many original roughness (sample variable), the
testing programmes is the standard of normal stress level and ultimately the
documentation. Obviously, the derivation condition of the final surface and debris
of samples should be recorded (ie, the pi'oduced.
borehole number and depth or field
location). However, the nature of Fig.7. Use of pro5le gauge to measure In multistage tests, the normal load is
28 individual samples before and after surface roughness of sample. increased (or decreased) after peak
GROUND ENGINEERING MARCH 1989
4
HG
200.
further lmm or so after peak to ensure that (1978); Hencher (1981) Gillette et al Natural joints through Dolomite,
shear strength is either falling or (1988)zz. Dokan, Iraq
remaining constant. After each stage, the 4 peak strengths. Data
In order to be able to analyse results uncorrected.
samples may be reset to their original
properly, horizontal and vertical MH.
position. However, this is often 4
displacements must be recorded 4
N
DOLOMITE
Shear and normal loads are converted to
available and the setting up of samples average 'engineering'tresses by
and proper documentation are the most dividing by the gross contact area of the
time-consuming parts of testing. 5OI. «NJME samples (this will vary throughout the
Rates of shearing below a few mm per test). The contact areas of ellipse shaped
minute do not affect test results as a rule. samples (from rock core) at given
horizontal displacements can be
0
0 5
HORIZONTAL
10
OISPLACEHEHT (oml
calculated using the following formula:
LEGEIG
44
FaLLING LOaO SERIES
HNJ 2
where A = gross area of contact
r-0
2a = length of ellipse
M6 HORIZONIAL 015PLACEHENT lmml
ln. 0:361
0:
I
2b = width of ellipse
/Oms
u = relative displacement
E
I
ZIH. CD= 266 Fig.9. Example of raw data.
0 TJL
100.
Block samples can often be approximated
typically recorded at set increments of as rectangular or other simple
0
horizontal displacement (say O. lmm). geometrical shapes for the purposes of
0
HORIZONTAL
5
OISPLACENENI (mml
10
However, closer intervals may be calculating stresses. The true contact area
required where, for example, the samples between samples is always much smaller
are dilating strongly with increasing shear than that calculated and the local stresses
LGG.
stress yet horizontal displacement is are correspondingly higher. However, the
essentially zero. The computer-based stresses calculated on the basis of the
3M STRESS CORRECTED
FOR DILATION
retrieval system in use at the University of gross or apparent contact area will be in
Leeds allows data from all channels to be accordance with those calculated for the
8 200
O
K
NORMAL STRESS lkH/mll On completion of a test, samples should displacement. It should be noted that
be examined, the nature of the surface whilst only single values of normal stress
Fig.8. Examples of results from rising damage described and the surfaces are given for each stage in Fig. 9
aud falling load multi stage tests. sketched and photographed. (corresponding to peak shear strength), 29
GROUND ENGINEERING MARCH . 1989
~
LEGESO
I
TCORRECKO
STRESSES CONNECTED
DATA
TNNI OILATICR
I
nm.
m
m EW lsn
the calculated normal stress does in fact EONAL SISESS (ks/mrl
dilation angles is evident in the
increasethroughouteachrundueto uncorrected plots for the four stages.
decreasing contact area. Fig.12.Stress path plots corrected and
In Fig. 14, the peak shear strength results
uncorrected for multi stage test (same
Fig. 10 shows results from two series of from Fig. 10 are replotted with stresses
test as Fig.9).
tests with peak shear stress plotted corrected for dilation. The reduction in
against normal stress. These show scatter is marked and the basic friction
considerable scatter and it is found that The dilational contribution to shear angle is clearly defmed by the corrected
the wider the range of roughness of strength can, therefore, be simply data points.
samples tested, the greater will be the accounted for by the geometry of the
scatter of this plot. Interpretation of results sample. Where shearing of major
presented in this way is extremely asperities accompanies dilation, a Influence of surface texture
difficult. An upper bound envelope, for cohesion component of shear strength will and, coating on results
example, would merely represent the be introduced but the contribution due to
strengths of the roughest samples tested Discontinuities from the same rock type
overriding can stQ1 be calculated on the
and the basic friction angle of the material and of similar surface texture and
basis of the geometrical path followed
could not be determined. mineralogy will generally yield a well
during shearing. The relative
defined strength envelope once
contributions to strength of basic friction
correction has been made for the effects
Correction of test data for and cohesion can be determined
of individual sample roughness. Friction
independently once the extra work done
dilation angles thus obtained will often be very
due to overriding has been accounted for.
different from the basic friction angle
As dilation occurs, work is done hi lifting
In practice, the incremental roughness measured either from tests on saw cut
the upper sample and this results in an
angle can be calculated throughout the surfaces of rock or from residual strength
increasing shear force being to re~ed test by considering the incremental tests. In tests on joints through Hong Kong
continue shearing. Patton (1966),
vertical (dv) and horizontal (dh) Granite (Fig. 13)the corrected value
conducting shear tests using toothed
displacements as follows: obtained for naturaQy textured surfaces
plaster models of simple geometry,'howed
was 6'igher than that obtained from tests
that the contribution due to =
dv/dh tan i on saw cut surfaces of slightly
dilation could be explained by the
decomposed granite (Hencher &
following relationship:
Stresses measured in the horizontal and Richards, 1982)'. The corrected friction
vertical planes can then be resolved angle for natural, uncoated joints through
r = 0 tan(8+ i) tangentially and normally to the plane dolomite was 10'igher than that obtained
along which shearing is actually taking from saw cut surfaces.
where r is shear strength, o is normal
place using the equations given in Fig. 11.
8
stress, is the basic friction of the
These corrected stresses may then be
Surface coatings may result in corrected
surfaces and i is the angle of inclination of friction angles considerably lower than
plotted to give a strength envelope
the toothed asperities causing dilation. those for clean discontinuities and
reflecting the shear strength of non
sometimes lower than those for saw cut
dilating, naturally textured surfaces. This
surfaces. Dolomite coated with minor
envelope may indicate purely frictional
amounts of bitumen and/or calcite was
behaviour (through the origin) or may
found to have a basic friction angle of
REASURED RORRAC
show additional cohesion which might be
STRESS lmmmm I - 0
) dv about 33'ompared to 39'or uncoated
expected to vary for different samples and
joints. Tests camed out on joints from
stress levels. In the experience of the
monzonite in Hong Kong gave values of
authors, the majority of persistent rock
joints exhibit essentially frictional
behaviour once dilation has been
accounted for.
In Fig. 12, shear stress is plotted against
Ettttf
+ ~
CINRECIEO
DATA
DATA
As shearing takes place, top block dilates with work normal stress (graduaQy increasing due to 1.0
being done against normal load through dv decreasing surface area in the case of
To calculate stresses relative to actual plane of shearing
uncorrected data) for four stages of a
direct shear test. Both uncorrected data
dv
Then, i = A Tan~ and data corrected for dilation are plotted
5; = (r cosi
—osini) cosi in this graph. A strength envelope can be
o;= (ocosi+ rsini)cosi derived directly from the corrected data.
In Fig. 13, the ratio between shear stress 5 u
Where compression takes place
then, r, = (r cosi+ osini) cosi
and normal stress is plotted versus EOEIEOEIAL OISPLACEREEI Immi
E
NN
(8
I NO.
I llg
0
El
ESNAL SN(SSS ACNNN SAm EA(s O; (0(NO(i termmed from tests on saw cut surfaces
orresiduaistren~
Determination of eff ~ve
Geld shear
stren requires detailed assessment of
c acteristics of dis
Some aspects of thts'rocedure have been
y 'edinCowland &
'chards (1982)s and will he furth
sequ p pe y
N TOOITC COAEO WITH
ETIHEHICAICTE
0
4 SNI NE
NINA( SNSSS, 0;
EN
(NN/ol)
References
1.Brown ET (1981)—Su e
'
detenninin gshearstrengtkRockchara
sh
Fig.14.Peak shear stzen stm and tut -ISRM
Fig.10 corrected for dilation. Pergamon Press, 138-137 influence of frost
The infl
38'or iron oxide stained joints b
a..ange Mand Herget G(19 -Det ernunauon of
discontinuities by direct
heave on houses under
where the rock had a thin 't
ope manual supplement 3-2: construction in Merth
port
tory tests for design
TydK %(Irwin