Author's personal copy

Engineering Geology 147-148 (2012) 91–100

Contents lists available at SciVerse ScienceDirect

Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo

Rock strength determination from scratch tests

Thomas Richard a, b, Fabrice Dagrain c, Edmond Poyol d, Emmanuel Detournay b, e,⁎
a
Epslog, Belgium
b
Earth Science and Resource Engineering, CSIRO, Australia
c
Département de Génie Civil et Mécanique des Structures de la Faculté Polytechnique de Mons, Université de Mons-Hainaut, Belgium
d
Rock Mechanics Laboratory, Total, France
e
Department of Civil Engineering, University of Minnesota, USA

a r t i c l e i n f o a b s t r a c t

Article history: This paper provides compelling experimental evidence that the unconfined compressive strength of rocks
Received 4 August 2011 can be reliably assessed from scratch tests performed with a sharp cutter, and at depth of cut small enough
Received in revised form 16 July 2012 to prevent any significant chipping of the rock. The paper describes the model used to interpret the experi-
Accepted 24 July 2012
mental results, the test methodology, and the Rock Strength Device that was developed to perform scratch
Available online 1 August 2012
tests under kinematically controlled conditions. It concludes with a summary of an extensive experimental
Keywords:
campaign involving the testing of several hundred rocks to compare strength data from conventional uniaxial
Uniaxial compressive strength compression experiments and from scratch tests.
Experiments © 2012 Elsevier B.V. All rights reserved.
Cutting test
Strength log

1. Introduction equipments used for sample preparation and testing have restricted
the test to a laboratory environment.
The uniaxial compressive strength (UCS) is the most common mea- Alternative indirect methods to estimate the UCS, such as the point
sure of strength used in civil, mining, and petroleum engineering, with load test, the indentation test, or the Schmidt hammer test, have been
applications ranging from the design of underground structures in rocks developed over the years (Broch and Franklin, 1972; Bieniawski, 1974,
to the selection of tools for mechanical excavation. The procedure to de- 1975; Chau and Wong, 1996; Szwedzicki, 1998; Rusnak and Mark,
termine the UCS has been standardized by the ASTM (2010) and the 1999; Aydin and Basu, 2005). Limited sample preparation, light equip-
ISRM (Ulusay and Hudson, 2007), and the test itself has been the subject ment, the partial or non-destructive nature of the testing are some of
of numerous publications (Bieniawski, 1968; Hawkes and Mellor, 1970; the benefits that these methods have over the uniaxial compression
Wawersik and Fairhurst, 1970; Broch and Franklin, 1972; Hudson et al., test. The main advantage of both the point load and indentation tests
1972; Jaeger et al., 1976; Pells, 1993). is actually their ability to assess strength with very small size samples
The uniaxial compression test used to determine the UCS suffers (about 3–5 cm3), while the main benefit of the Schmidt hammer test
from several drawbacks, however. It requires not only cores of intact is that it allows for direct testing on outcrops in the field.
rock but also precise and time consuming sample preparation, in partic- These indirect methods are not without problems, however. First,
ular the rectification and the polishing of the sample ends. The magni- there is documented subjectivity in the correlation between the point
tude of the axial stress at which the rock fails depends on the aspect load index or the Schmidt rebound number and the UCS (Kahraman,
ratio of the core, one of the reasons why the procedure for determining 2001; Fener et al., 2005). Second, both test results are affected by the
the UCS has been standardized. The moisture content and the irregular elastic properties of the material, the sample size and the moisture con-
ends are also known to affect the test outcome (Bieniawski, 1968; tent (Bieniawski, 1975; Tsur-Lavie and Denekamp, 1982; Kahraman,
Hawkes and Mellor, 1970; Broch and Franklin, 1972; Hudson et al., 2001; Thuro et al., 2001; Tsiambaos and Sabatakakis, 2004; Aydin and
1972; Hoek and Brown, 1980; Dey and Halleck, 1981; Farmer, 1992). Basu, 2005). There are other issues, such as the point load testing of ir-
Moreover, the test is not well suited for heterogeneous, damaged, regular shaped material (Broch and Franklin, 1972), or the roughness or
layered or fractured rocks. In such materials, failure is often controlled degree of polishing of the surface on which the impact occurs, which
by the weakest plane, joint or a pre-existing crack present in the core also affect the rebound values (Hucka, 1965).
sample. Finally, the requirements in terms of power and size for the The scope of this paper stems from an effort initiated at the Uni-
versity of Minnesota (UMN) in the mid-nineties to build a scientific
⁎ Corresponding author at: Department of Civil Engineering, 500 Pillsbury Drive SE,
apparatus to study the cutting action of a single cutter (Detournay
Minneapolis, MN 55455, USA. et al., 1997) in order to assess the dependence of the cutting force on
E-mail address: [email protected] (E. Detournay). the rock mechanical properties and on the UCS, in particular (Almenara

0013-7952/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.enggeo.2012.07.011
 Author's personal copy

92 T. Richard et al. / Engineering Geology 147-148 (2012) 91–100

and Detournay, 1992; Detournay et al., 1995; Adachi et al., 1996; Richard is based on three key assumptions applicable to a particular cutter–
et al., 1998). Comparable efforts by other researchers aiming at using re- rock combination, irrespective of the cutter wear, namely (i) the forces
sults of both cutting and drilling tests to estimate rock strength have also on the cutting face, suitably averaged over a distance large compared to
been recently documented (Reddish and Yassar, 1996; Perrier, 1997; the depth of cut, is proportional to the cross-sectional area Ac of the
Tiano, 2001; Balci et al., 2004; Stavropoulou, 2006; Pamplona et al., 2007). groove traced by the cutter; (ii) the inclination of the average force on
In this paper, we first review basic aspects of the mechanics of rock the cutting face is constant; and (iii) there is frictional contact at the
cutting, with considerations given to the ductile and the brittle regimes, wear-flat rock interface. Such a model is characterized by three param-
and to the influence of a wear flat on the magnitude of the cutting force. eters: the intrinsic specific energy ε associated with the cutting process,
We then describe the Rock Strength Device, as well as the test method- the number ζ giving the inclination of the force acting on the cutting
ology. Finally, we produce compelling evidence that under conditions face, and the friction coefficient μ mobilized across the wear flat.
referred to as the ductile regime, the specific energy of cutting is well The force F acting on a blunt cutter results therefore from the super-
correlated with the UCS. position of two forces Fc and Ff, acting on the cutting face and on the
wear flat, respectively. The components of these two forces in a direc-
2. Mechanics of rock cutting tion parallel (subscript s) and perpendicular (subscript n) to the cutter
motion can thus be expressed as (Figure 1)
A scratch test involves tracing a groove on the surface of a specimen
with a cutting tool. The test has been widely used (Nishimatsu, 1972; F cs ¼ εwd; F cn ¼ ζ εwd; ð1Þ
Duc, 1974; Deliac, 1986; Glowka, 1989b) to study the cutting process
and to assess the effect of rock properties and tool characteristics on where a rectangular cutter has been assumed, i.e., Ac = wd, while the
the performance of drilling or excavation tools. components of the force Ff acting on the wear flat are constrained by
The scratch test is typically conducted under kinematic control: the
depth of the cut d (or depth of the groove) and the cutter velocity v F fn ¼ μF fs : ð2Þ
(tangential to the sample surface) are imposed and maintained con-
stant along the entire cut, while the magnitude and orientation of the The intrinsic specific energy ε represents the energy required to cut
force acting on the cutter are measured, see Figure 1. a unit volume of rock. The word “intrinsic” emphasizes that this energy
The cutter–rock interaction is generally characterized by the coexis- is strictly used to cut the rock; in other words, ε does not take into ac-
tence of two processes, namely rock fragmentation in front of the cut- count frictional dissipation associated with cutter bluntness. Thus, “in-
ting face of the tool and frictional contact along the wear flat/rock trinsic” does not imply that ε is independent of the cutter geometry
interface (Fairhurst and Lacabanne, 1957; Duc, 1974; Glowka, 1989a; (shape and back-rake angle) or even the depth of cut. However, the
Detournay and Defourny, 1992). Moreover, there exist two different phenomenological model, on which the interpretation of scratch data
cutting mechanisms, ductile and brittle, whose occurrence is controlled is based, assumes that ε is indeed independent of the depth of cut and
by the magnitude of the depth of cut (Chaput, 1992; Richard et al., 1998; it is a constant quantity characterizing a particular combination of cut-
Richard, 1999; Huang and Detournay, 2008; Huang et al., 2012). ter geometry and rock. The intrinsic specific energy is expressed here
At shallow depth of cut (typically less than 1 mm for a medium as a stress (e.g., in unit MPa), rather than as an energy per volume
strength sandstone), the rock is intensively sheared ahead of the cut- (e.g., in unit J/cm3, which is numerically equivalent to MPa), on account
ter; this cutting regime is mainly characterized by a de-cohesion of of the expected relation between ε and the UCS q.
the constitutive matrix and grains, with grains and powder accumu- The number ζ can be expressed as tan(θ + ψ) with ψ the inclination
lating progressively in front of the cutter (Figure 2). angle of the force to the normal to the cutting face, see Richard (1999);
At larger depth of cut, brittle failure occurs, with macroscopic Huang et al. (1999); Coudyzer and Richard (2005) for more details on
cracks initiating from the tool tip and propagating unstably ahead of the meaning of this angle. For a given back-rake angle, the angle ψ is
the cutter; chips, fragments of rock are created and removed by the found to be independent of the depth of cut.
cutter (Figure 2). Finally, μ is the coefficient of friction on the wear flat/rock interface,
As expected, experimental results indicate that different material which can also be expressed as μ = tanϕ where ϕ is the angle between
properties control the magnitude of the cutting force in the two the frictional force Ff and the normal to the wear flat.
modes of failure (Richard et al., 1998): fracture toughness KIc in the Combining the two processes (pure cutting and frictional contact)
brittle regime and the uniaxial compressive strength, q in the ductile leads to a relationship between the cutting Fs and normal Fn compo-
regime. The depth of cut characterizing the transition between ductile nents of the total force
and brittle regime scales by the rock intrinsic length scale (KIc/q) 2
F s ¼ εð1−μζ Þwd þ F n ; ð3Þ
(Huang and Detournay, 2008; Huang et al., 2012). These two mechan-
ical properties reflect different modes of energy dissipation: along
which scaled by the cross-section area Ac = wd yields a constraint on
created macroscopic discontinuity surfaces in the brittle regime and
the cutter response
within the volume of the failed material in the ductile regime.
A phenomenological model of cutter/rock interaction in the ductile
E ¼ Eo þ μS; ð4Þ
regime was proposed by Detournay and Defourny (1992). This model
where E and S are the specific energy and the drilling strength respec-
tively defined as
Fc
Fs Fn
v E¼ S¼ ð5Þ
wd wd
; ;
Ff
s and Eo = ε(1− μζ). Accordingly, points pertaining to tests performed
n d with a blunt cutter define the friction line in the ES diagram. The posi-
tion along the friction line is controlled by the cutter state of wear (mea-
sured as the length ℓ of the wear flat in the case of rectangular cutter),
the depth of cut, as well as the normal stress σ on the wear flat. A syn-
Fig. 1. Cutting configuration. Forces acting on a blunt cutter. thetic ES diagram is shown in Figure 3.
 Author's personal copy

T. Richard et al. / Engineering Geology 147-148 (2012) 91–100 93

Fs (N) Fs (N)
160 600
a b
80 300

0 0
0 20 40 60 0 20 40 60
s (mm) s (mm)

Fig. 2. Schematic of the ductile and brittle regimes. Examples of force signal (tangential component Fs) in the ductile (d=0.6 mm) and brittle regime (d=3 mm). Tests conducted in a
specimen of Vosges sandstone.

It is important to reiterate the difference between the intrinsic spe- The above experimental evidence supports the assumptions behind
cific energy ε and the specific energy E. The latter quantity, E, accounts the model of rock cutting in the ductile regime, namely that the param-
for both the energy expended in cutting the rock and the energy dissi- eters ε, μ, and ζ are constant for a particular rock/cutter pair, irrespective
pated at the frictional contact between the wear flat and the rock, of the cutter wear. However, the inclination of the cutting force appears
while the former one, ε, only characterizes the energy used to fragment to be weakly dependent on the rock and to only vary in a narrow range.
the rock ahead of the cutter. Indeed, the histogram of ζ, measured in scratch tests conducted on
Figure 4 shows evidence of the proportionality between the magni- about 375 rocks (Appendix A), using a cutter with a back-rake angle
tude of the cutting force and the depth of cut, for scratch tests of 15°, shows that ζ is distributed normally with a mean of 0.6 and a
conducted in the ductile regime with a sharp cutter. The rocks used in standard deviation of 0.16. This relative invariance of ζ can actually be
the tests have been selected for their high degree of homogeneity; used for monitoring the wear state of the cutter, as values of ζ larger
they consist of two sandstones (Vosges and Rhune), three limestones than 0.9 suggests the existence of a force on the wear flat that is not neg-
(Anstrude, Buxy, Fontenoille) and one chalk (Harmignies). For each ligible compared to the cutting force (Figure 6). Although nothing has
rock, a best linear fit is conducted on the set of data points (Fs, d). been said on the magnitude of the force Ff at the wear flat/rock contact,
Using Eq. (1), the intrinsic specific energy is then simply inferred as there is evidence that the mean normal contact stress Ffn/Afn, where Afn
the estimated slope divided by the cutter width, w. Tables A1–A3 in is the nominal wear flat area, although sensitive to the relative inclina-
Appendix A list the UCS q and the intrinsic specific energy ε for the six tion of the cutter velocity on the wear surface, is bounded by the UCS
rocks of Figure 4, as well as for many others. (Almenara and Detournay, 1992; Adachi et al., 1996).
An example of a ES diagram with points pertaining to tests performed
with a blunt cutter on four rocks (shale, coal, sandstone, and limestone) is 3. Rock Strength Device
shown in Figure 5. The friction coefficient μ can be inferred as the
slope of the friction line (estimated from best linear fits conducted The Rock Strength Device (RSD) is a testing apparatus that was de-
on the data set). The estimated values of μ and ϕ for the four rocks veloped at the University of Minnesota in the late 1990's (Detournay
are listed in Table 1. These results are consistent with previous ob- et al., 1997). The RSD scratches the surface of rock samples under pre-
servations that the friction angle ϕ characterizing the cutter wear cise kinematic control, while enabling accurate measurements of the
flat/rock contact is often reminiscent of the internal friction angle force acting on the cutter. The main components of the frame (see
of the rock (Almenara and Detournay, 1992; Detournay and Figure 7) are: a traverse with a sample holder (indexed 1 on Figure 7),
Defourny, 1992; Dagrain and Richard, 2006).

E 1000
Harmignies chalk
friction line Vosges sandstone
800 Anstrude limestone
1 Charmot limestone
Fontenoille sandstone
ζ Rhune sandstone
µ 600
Fs (N)

ε 1
Eo 400
cutting point

200

S 0
0.2 0.4 0.6 0.8 1.0 1.2
Fig. 3. Synthetic ES diagram (Detournay and Defourny, 1992). The position of the state d (mm)
point on the friction line depends on the cutter bluntness and depth of cut. For a per-
fectly sharp cutter operating in the ductile regime, E = ε. The cutting efficiency ε/E de- Fig. 4. Variation of the tangential component of the measured force with the depth of
creases with depth of cut d and/or with increasing wear flat length ℓ. cut d. Tests performed with a sharp rectangular cutter of width of w = 10 mm.
 Author's personal copy

94 T. Richard et al. / Engineering Geology 147-148 (2012) 91–100

400 60
Boom shale
Coal
Vosges sandstone
Lens limestone
40
300
E (MPa)

20
200

0
0.0 0.2 0.4 0.6 0.8 1.0
100
ζ

Fig. 6. Histogram of ζ measured for about 300 rocks, see Appendix A.

0
0 100 200 300 400
the results for w = 2.5 mm and 5 mm, where ε is seen to increases with
S (MPa)
d; (ii) the intrinsic specific energy ε hardly varies when comparing the
Fig. 5. ES diagram for four different rock materials. Tests conducted with a blunt cutter. results for w = 10, 15, and 50 mm.
The variation of ε with w and d that is observed for w = 2.56 and
5 mm is thought to be related to the existence of another length
a moving cart (2) housing the vertical positioning system (3), the load scale, the grain size, and to the dilatant nature of shear failure. In nar-
cell (4), and the cutting element (5). The horizontal movement of the row grooves, the shear-dilatant behavior of the rock could lead to a
cart is operated by a computer controlled stepper-motor (6) driving transverse normal stress (perpendicular to the cutting direction) as
an horizontal ball screw (7) via a gearbox (8). The depth of cut is adjust- the chain of grains are short enough to offer some resistance to
ed manually with the vertical positioning system (9) and a micrometer “buckling.” (Vosges sandstone, the average grain diameter is about
(10). A locking system (11) secures the vertical traveling mechanism 0.15 μm, meaning that only 10 to 20 grains fill the groove width for
against the frame, in order to maintain a constant depth of cut while w = 2.56 mm.) Furthermore, it is also expected that the stability of
cutting. the granular chains increases with the depth of cut. The frictional
A load sensor measures the tangential and normal components (Fs, strengthening associated with the transverse normal stress is then
Fn) of the force F acting on the cutter. The complete system (sensor, translated into an increase of the specific energy.
data acquisition) achieves about 1 N of precision and resolution over The depth of cut (ranging typically between 0.1 and 1 mm) is the
the entire measurement range [0–4000 N]. The scanning rate is typical- only parameter controlling the magnitude of the force acting on a cut-
ly set at 25 samples per millimeter travelled by the cutter. ter, given its geometry and the rock being scratched. As the test is
Two types of cutting tools are used: sharp and blunt cutters. Sharp performed under imposed depth of cut, precise adjustment of the
tools present only one contact surface with the rock, the cutting face, depth of cut and minimal compliance of the frame are thus critical.
which is inclined forward by the back-rake angle θ, here set to 15°. In The vertical compliance of the frame/sensor assembly was estimated
addition to the cutting face, blunt cutters possess a machined “wear by measuring the vertical displacement of the cutter tip under an im-
flat,” moving parallel to the cutting direction against the bottom of the posed axial force; it was found equal to about 0.038 mm/kN for the
groove. The flat surface is inclined forward by a small angle (β ~ 2°) to Rock Strength Device depicted in Figure 7.
ensure conforming contact with the rock (Detournay et al., 1997). For An experiment was designed to assess the reliability of the depth of
rectangular shaped cutters, the geometry chosen for the tests, descrip- cut adjustment; the nominal depth of cut dn set prior to the test with the
tion of the cutter geometry reduces to the width w (w = 10 mm) and micrometer, was compared to the mean effective depth of cut de mea-
the wear flat length ℓ (varying from a fraction of mm to a couple of sured with a laser as the thickness of material removed during the
mm). The cutting face of the cutters commonly used are made of a test. The vertical distance between the centre line of the groove bottom
thin layer of polycrystalline diamond compact (PDC) laid down on a surface and the same absolute reference point is logged using a laser,
carbide tungsten base. before and after the cut, thus providing the variation of the effective
A campaign of scratch tests have confirmed that the intrinsic specific depth of cut along the test. The results for a medium strength rock
energy is indeed relatively insensitive to the cutter width w and the (q~ 50 MPa) show that the absolute difference between dn and de falls
depth of cut d for width around w = 10 mm and beyond, if cutting within the precision of the micrometer, see Figure 9.
takes place in the ductile regime. Figure 8 shows results of tests In addition, the overall stiffness of the assembly (frame and sensor)
conducted using five cutter widths (w = 2.56,5,10,15, and 50 mm) in is sufficiently large to ensure that its dominant natural frequencies
Vosges sandstone at depths of cut ranging from 0.2 to 1.8 mm (around 600–700 Hz for the sensor) are far outside the bandwidth of
(Richard, 1999; Dagrain, 2006). The figure reveals that (i) the specific the cutting force signal (about 98% of the energy associated to the signal
energy for w = 10, 15, and 50 mm indeed does not depend on d, unlike sits between 0 and 10 Hz).

4. Test methodology

Table 1
The testing procedure consists in performing successive scratch
Friction coefficient μ and friction angle ϕ.
tests along the same groove on the surface of a rock specimen, while
Rock μ ϕ measuring the force components Fs and Fn, which are respectively par-
Boom shale 0.22 12.4 allel and normal to the cutter velocity. Each test is conducted along the
Coal 0.39 21 entire specimen length under constant depth of cut d and constant cut-
Vosges sandstone 0.63 32.4
ting velocity v. The tests reported in this paper were carried out with
Lens limestone 0.82 39.3
cutters characterized by a width w = 10 mm and a back rake angle
 Author's personal copy

T. Richard et al. / Engineering Geology 147-148 (2012) 91–100 95

Fig. 7. Sketch of the Rock Strength Device (front and side view).

θ = 15° and with the cutter velocity set to v = 4 mm/s. The depth of cut (Figure 10). This result also suggests that within the ductile regime, the
d typically ranged between 0.1 and 1 mm, with increment varying from mean force is the only relevant quantity.
0.05 to 0.3 mm between successive cuts. The force components are The question of the appropriate length scale over which the signal can
measured at a sampling rate of 100 Hz, meaning 25 measurements of be meaningfully observed (i.e., averaged) is important for interpreting the
force per mm of cut. force results. One simple criteria concerns the reproducibility of the test
The tests can be carried out using either sharp or blunt cutters, as results. The relevant and reproducible information, in terms of mechani-
discussed below. For tests with sharp cutters, the sharpness of the cutter cal properties, is evidently associated with the background trend of the
needs to be constantly monitored. Besides visual inspection of the cut- force signal, at “low” spatial frequencies, while the “high” frequencies
ter tip between each test series, the sharpness is most easily monitored events are not reproducible because they are associated with stochastic
by tracking the evolution of the number ρ= Fn/Fs between each cut. The events such as dislodgement of individual grains. Experience suggests
force inclination is indeed a useful indicator for evaluating the cutter that the force signal should be averaged over a length scale of a few
sharpness, as wear causes an increase of the ratio ρ of the normal to tan- millimeters to provide repeatable mean value of the force signal in
gential component of the total force acting on the cutter (ρ = ζ for an homogeneous sections. Fig. 11 illustrates the variation of the cutting
ideally sharp cutter). For tests with blunt cutters, consideration must force recorded while scratching the surface of a heterogeneous shale
be given to the wear flat size, so as to limit the forces mobilized by the sample (d = 0.4 mm); it shows the existence of four distinct regions,
blunt tool to levels acceptable for the sensor. each characterized by an approximate “stationary response” and
In order to estimate the rock strength q from cutting tests, the tests thus by different mean strength properties.
have to be conducted within the ductile regime. Various indicators, such
as mechanical vibrations, noise, and size of debris, exist to ascertain 5. Strength determination
whether cutting takes place within the ductile or the brittle regime. In
addition, the force signal itself can be used for the purpose, as its prop- An extensive campaign of testing to compare results from uniaxial
erties differ in the ductile and in the brittle regime. In the ductile regime, compression tests and scratch tests has been ongoing since 1995 in sev-
the force signal may be viewed as a white noise, whereas the signal in eral rock mechanics laboratories (University of Minnesota, USA; Faculté
the chipping mode presents a marked sawtooth pattern (Figure 2). Polytechnique de Mons, Belgium; Total, France). Results for 376 differ-
The coefficient of variation γ, the ratio of the standard deviation to the ent rocks are reported in this paper (Appendix A); these rocks consist
mean, was found to be a good indicator of the difference between the of 130 quarry materials (96 limestones, 25 sandstones, 2 chalks, 2
two types of signals. Indeed, γ is almost independent of the depth of shales, 3 granites and 2 schistes) and 246 non referenced rocks, mainly
cut within the ductile regime, but increases with d in the brittle regime
1

100
w = 2.56 mm
5.05 mm
10 mm 0.75
80
15 mm
50 mm

60
ε (MPa)

de (mm)

0.5
40

20
0.25
0
40 80 120
w/d

Fig. 8. Variation of intrinsic specific energy ε with w/d for tests conducted on Vosges 0
0 0.25 0.5 0.75 1
sandstones with rectangular cutters with various widths (w = 2.5, 5, 10, 15, and
50 mm). The depth of cut varies as follows: w = 2.56 mm, d = [0.2,1] mm; w = 5 mm, dn (mm)
d = [0.3,1.5] mm; w = 10 mm, d = [0.2,1.8] mm; w = 15 mm, d = [0.3,1.4] mm; w =
50 mm, d = [0.5,0.8] mm (Richard, 1999; Dagrain, 2006). Fig. 9. Comparison between the nominal dn and effective de depths of cut.
 Author's personal copy

96 T. Richard et al. / Engineering Geology 147-148 (2012) 91–100

0.6 250
limestones
sandstones
others

200
0.4

γ
150

ε (MPa)
0.2

100
0
0 0.5 1 1.5 2
d (mm) 50

Fig. 10. Evolution of the coefficient of variation γ of the tangential force component
with the depth of cut d.
0
0 50 100 150 200 250

core samples from oil or gas reservoirs (310 sandstones, 10 limestones q (MPa)
or dolomites and 24 shales or pelites, 2 granites).
Fig. 12. Correlation between the intrinsic specific energy ε and the uniaxial compres-
The correlation between the intrinsic specific energy ε and the UCS q is
sive strength q. The data of Appendix A have been grouped into limestones, sandstones,
shown in Figure 12. The data pertaining to the quarry materials are sum- and others.
marized by rock type in Tables A1–A3, which are given in Appendix A.
A few points have been eliminated from the correlation plot because
the results of the uniaxial compression were considered as not being energy estimated at each location. Eight samples were then cored out of
representative of the material strength. As an example, samples of a the block with the axis of the cores parallel to the cutting direction, and
shale provided scattered results for q varying from 2 to 26 MPa as the subjected to uniaxial compression.
samples were failing along fractures present in the specimen. Cutting The results, shown in Table 2, highlight a dispersion among the com-
tests performed on the same samples were not affected by the fracture pression test results that is about one order of magnitude higher than
planes or de-lamination planes (ε = 71 MPa). for the scratch results; i.e. a coefficient of variation (standard deviation
to mean) of 1.3% for the scratch test against 13% for the uniaxial com-
6. Benefits of the test pression test. Furthermore, tests carried out on core samples of Lens
limestone with diameters ranging from 19 to 69 mm show a dispersion
One essential feature of the scratch test is its high degree of repeat- in the force measurement comparable to the dispersion observed in re-
ability. Results of tests carried out on an homogeneous rock sample sults of tests conducted on the same sample (Figure 14). The inferred
display little dispersion, as illustrated in Figure 13. For this example, a values of the intrinsic specific energy have a mean of 23.4 MPa with
coefficient of variation of about 4% is obtained for the two force compo- standard deviation of 0.8 MPa, corresponding therefore to a coefficient
nents from 30 similar scratch tests conducted with a sharp cutter on a of variation of about 3%. Similar results are summarized in Table 3 for
sample of Indiana limestone at constant depth of cut (d = 0.2 mm) tests performed on blocks of Vosges sandstone with thickness varying
over a distance of 130 mm. from 100 mm down to 6 mm.
Scratch as well as compression tests were performed on samples of It is also interesting to note that results of tests conducted on a saturat-
Vosges sandstone to compare the dispersion of the strength measured ed sample sit within the error margin of the results obtained on dry sam-
from these two tests. A rectangular block (200× 200 × 100 mm) of ples (Figure 14), suggesting that the moisture content has negligible effect
Vosges sandstone was prepared and scratch tests performed at 10 dif- on the test outcome, at least for rocks whose structure is not altered by
ferent locations on one outer surface of the sample; all the cuts were the degree of saturation. For example, there is evidence that the intrinsic
in the same direction. For each location, five scratches were carried specific energy on some shales is affected by hydro-chemical interactions.
out at depth of cut ranging from 0.1 to 0.5 mm and the intrinsic specific The above results indicate that the intrinsic specific energy can be
estimated with confidence from the result of a single scratch performed

600 200

450 150
Fs
Fs , Fn (N)
Fs (N)

300
100

Fn
150
50

0
0 20 40 60 80 0
s (mm) 0 8 16 24 32
Test Number
Fig. 11. Example of a force signal (tangential component) characterized by 4 distinct
regions. Tests performed at d = 0.4 mm. A low-pass filter with a cut-off frequency cor- Fig. 13. Averaged force components (Fs, Fn) for a series of 30 scratch tests performed on
responding to 0.25 mm has been applied to the raw force signal. a sample of Indiana limestone (at d = 0.2 mm) over a length of 130 mm.
 Author's personal copy

T. Richard et al. / Engineering Geology 147-148 (2012) 91–100 97

Table 2 Table 3
Dispersion in the results of uniaxial compressive and scratch tests, performed on samples Dispersion in the results of scratch tests performed on samples of Vosges
of Vosges sandstone. (For the UCS tests: coefficient of variation γ(q)=13%, corresponding sandstone characterized by different thickness. (Coefficient of variation
to a standard deviation of 6 MPa and a mean of 45.7 MPa. For the scratch tests: coefficient γ(ε)=2.7%, corresponding to a standard deviation of 1.2 MPa and a mean
of variation γ(ε)=1.3%, corresponding to a standard deviation of 0.7 MPa and a mean of of 45.6 MPa.)
50.3 MPa.)
Sample thickness (mm) ε (MPa)
Test number q (MPa) Test number ε (MPa)
100 44.9
1 51.6 1 50.4 100 44.2
2 46.8 2 51.1 25.5 45
3 40 3 49.9 18.5 45.2
4 45.2 4 50.9 14 46.4
5 48.3 5 50 11 48.1
6 34 6 51 8 46
7 49.1 7 49.9 6 44.8
8 50.6 8 49
9 – 9 50.7
10 – 10 50.6
The scratch test offers several advantages over conventional tests.
First, the results are not affected by the sample dimension, and thus
at a depth of cut small enough to ensure dominance of the ductile re- only a small volume of intact rock (10 to 20 mm thick, wide and
gime and large enough to neglect the effect of the cutter bluntness (de- long) is required to assess its strength. Second, the sample prepara-
fects along the cutting edge). Although E = Fs/wd≃ ε for very sharp tools tion is limited as it only requires performing a pre-cut on the sample
(ℓ b 0.05 mm), it is preferred, however, to perform additional scratches surface to obtain a flat reference surface that ensures a constant depth
at different depths of cut and extract ε from the slope of the linear re- of cut along the groove; no additional equipment is thus needed for
gression line calculated from the pairs (Fs, d). sample preparation. Third, the test offers a high degree of repeatabil-
In summary, the intrinsic specific energy appears to be unaffected by ity quite uncommon in strength testing of geomaterials. Fourth, the
the sample dimensions as long as a few criteria are satisfied. First, the semi-destructive nature of the test is an interesting asset, as it allows
scratch length should be equal to at least a few millimeters. Second, additional group of tests to be conducted on the same sample (poros-
the sample width should be sufficient to accommodate a cutter of ity, permeability, sound velocity, uniaxial compression, triaxial test,
10 mm width. Finally, the sample thickness should be larger than the etc…). Fifth, the scratch technique naturally yields a log of the rock
depth of cut by a few millimeters. strength, in contrast to the direct and indirect methods which only
The continuous nature of the scratch test associated to a spatial res- yield punctual measurements of the strength; the information about
olution of order of millimeters naturally yields a log of the material the spatial variation of the strength provides not only invaluable in-
strength. This strength log can be approximated by the log of the intrin- sights on the inhomogeneity of the rock but also a context to the con-
sic specific energy E by simply dividing the filtered force signal by the ventional meaning of the UCS. Finally, the equipment is portable, can
product wd (Detournay et al., 1995; Dagrain et al., 2004). Practically, be manipulated on site, and the test procedure is quick as parameters
the method allows capturing variations of the order of a few MPa over are obtained within a few minutes. The scratch test is thus an attrac-
a distance of a few millimeters along the sample surface. Logs of the spe- tive and elegant alternative to conventional uniaxial compression
cific energy E measured at different depth of cut from tests performed tests and to indirect methods such as the point load test and the
on the surface of an heterogeneous reservoir rock are shown in Figure Schmidt hammer test.
15, and show evidence of the reproducibility and resolution of the
scratch technique. Acknowledgment

7. Conclusions Many people have contributed to the development of the scratch

test to measure the strength of rocks and to the design of the RSD.
This paper has provided compelling experimental evidence that the Their contributions are gratefully acknowledged, with the authors
unconfined compressive strength of rocks can be reliably assessed from being particularly indebted towards Robert Delwiche, Andrew
scratch tests performed under controlled conditions, namely with a Drescher, Erling Fjaer, David Hultman, Cécile Lasserre, and Tanguy
sharp cutter and at depth of cut small enough to guarantee that cutting Lhomme for their help. The research was funded by grants from the
takes place in the ductile regime. Sota Tec Fund, Elf Aquitaine, Total, and DBS.

300 60
D = 19.4 mm d = 0.3 mm
24.5 mm d = 0.4 mm
33.5 mm d = 0.5 mm
225 39.6 mm
60.2 mm
69.1 mm 40
E (MPa)

24.8 mm - saturated
Fs (N)

150

20
75

0 0
0 3 6 9 12 0 100 200 300
wd (mm2) s (mm)

Fig. 14. Effect of size and moisture on the response of a sharp cutter, tests performed on Fig. 15. Log of the specific energy measured at different depth of cut on the surface of
cores of Lens limestone characterized by different diameters D. heterogeneous limestone.
 Author's personal copy

98 T. Richard et al. / Engineering Geology 147-148 (2012) 91–100

Appendix A. Compilation of measured specific energy and strength (continued)

Table A.1 (continued)
data Rock q (MPa) ε (MPa)

Fontenoille 87.0 88.0

Rouge des Flandres 108.6 88.0
Table A.1 Balzac Fleuri 90.0 90.5
Carrare 50.9 90.6
Intrinsic specific energy ε and uniaxial compressive strength q of quarry limestone
Vinalmont 84.6 92.0
materials.
Chassagne Beauharnais Beige 110.0 95.3
Rock q (MPa) ε (MPa) Moleanos 96.0 96.2
Carrare 57.7 97.1
Saint Vaast 6.0 5.8 Comblanchien 85.5 97.9
Saint Maximin Ferme Const. 9.0 7.2 Soignies 100.9 98.0
Pombreton 7.0 7.3 Chassagne Beauharnais Beige 110.0 98.4
Nurabup 8 9 Gris Catalan 130.0 107.1
Sireuil Dore 8.0 9.3 Carrare 112.0 107.6
Sireuil Beige 11.0 11.1 Balzac 110.0 108.5
Richemont Jaune 13.0 13.2 Soignies 113.4 110.0
Lavoux 12.6 15.3 Comblanchien 127.0 113.3
Richemont Bleu 13.0 15.3 Gobertange 96.1 115.0
Tervoux 30.0 15.5 Neufvilles 115.3 118.0
Saint Maximin Franche Const. 20.0 16.1 Rocherons Legerement Mouchete 125.0 123.3
Saint Maximin Ferme Fine 15.0 18.3 Rocherons Dore 125.0 126.3
Savonniere 12.5 18.3 Crema Marfil 113.0 130.3
Estaillades 8.9 18.4 Pierre Bleue Belge 158.0 132.7
Saint Maximin Franche Const. 20.0 22.5 Crema Marfil Alicante 113 133.0
Chauvigny Roche Fine 25.0 23.7 Chassagne Beige 160.0 133.4
Grand Court 15.4 24.0 Buxy Bleu 105.0 137.3
Saint Remy Clair 48.0 25.8 Gris St Anne 160.0 159.0
Saint Nicolas Rubane 40.0 28.1 Asian Blue 169.0 161.9
Charmont 37.5 28.9 Marbre noir 193.0 218.8
Euville 28.0 31.6
Anstrude Roche Claire 40.0 31.7
Saint Nicolas a Contre-Passe 40.0 32.6 Table A.2
Chauvigny Classique 40.0 33.4
Anstrude 22.6 34.0 Intrinsic specific energy ε and uniaxial compressive strength q of quarry sandstone
Rosal 26.4 35.0 materials.
Euville Metropolitain 28.0 35.3
Rock q (MPa) ε (MPa)
Chauvigny 29.0 36.0
Lens 45.0 38.3 Mc Field 3.8 3.4
Indiana 17.1 40.7 Salt Wash South 7 8.6
Massangis Roche Claire Nuancee 50.0 41.1 Redwilmore 12.1 12.6
Valreuil Perle 45.0 41.2 Castelgate 18 16.6
Lens 38.4 42.0 Redwilmore 13.0 17.0
Boisvillon 32.7 42.1 Castelgate 15.0 22.0
Tuscany 87.0 43.0 Vosges Rouge 13.5 24.6
Vilhonneur Classique 75.0 43.0 Vosges Rouge 15.9 27.0
Vilhonneur Bercy 40.0 43.1 Vosges Vert 17.2 27.2
Anstrude 48.0 46.6 Vosges Rouge Bigarre 43.9 31.8
Massangis Roche Claire Nuancee 50.0 46.7 Vosges 42.0 32.0
Euville 29.8 48.0 Cleeburg 29.6 37.2
Valanges a Contre-Passe 40.0 48.5 Berea 42.6 41.5
Brousse Perle 70.0 49.1 Colton 43.8 46.9
Charmot 50.0 49.7 Annot 42.3 50.0
Carthage 54.0 50.5 Rhune 48.8 77.8
Moka 60.5 53.0 Fontainebleau 110.0 85.0
Buxy 49.1 54.4 Fontainebleau 85.3 87.3
Moka 54.8 54.7 Rhune 110.0 122.0
Vilhonneur 23.4 55.0 Rhune 159.0 132.7
Massangis Roche Jaune 65.0 56.6 Yvoir 133.1 158.0
Pouillenay 70.0 59.1 Ourthe 138.4 160.0
Tavel 75.6 59.6 Fontainebleau 178.0 160.3
Buxy 78.0 60.0 Chasse 183.5 175.0
Noir de Mazy 115.3 60.1 Lauze 171.0 223.0
Indiana 28.0 60.2
Fatima Claire 83.0 61.2
Massangis Roche Jaune 65.0 62.7 Table A.3
Bleu de Ligneres 100.0 63.0
Buxy 78.0 66.0 Intrinsic specific energy ε and uniaxial compressive strength q of other quarry materials.
Massing 37.9 68.5
Saint Corneille Blanc 90.0 69.1 Rock Type q (MPa) ε (MPa)
Travertin 84.0 69.6 Meudon Chalk 1.2 2.9
Rose Aurore 67.3 69.9 Harmignies Chalk 4.9 5.0
Massangis Jaune Claire Nuancee 60.0 70.0 Boom Shale 2.5 5
Simyra 102.0 71.4 Pierre Shale Shale 5 6
Sofia Creme 87.0 76.7 Schiste de Martelange (parallel to bedding) Slate 58.3 51.0
Villefort 90.0 79.3 Schiste de Martelange (orthogonal to bedding) Slate 126.4 145.0
Rose Aurore 90.0 84.0 Lac du Bonnet Granite 86.0 123.2
Vilhonneur Classique 75.0 86.5 Tarn Granite 144.1 125.0
Marlin Granite 173.0 165.5
 Author's personal copy

T. Richard et al. / Engineering Geology 147-148 (2012) 91–100 99

Table A.4

q (MPa) ε (MPa) q (MPa) ε (MPa) q (MPa) ε (MPa) q (MPa) ε (MPa)

7 2 13.1 18.3 19.2 32 51.2 51.9

7.5 2.7 27 18.7 46.1 32.2 45.9 52
13.2 3.4 20 18.7 46 32.3 53.9 52.4
1.6 4.3 11.2 19.4 33.6 32.4 42.8 52.9
22 4.3 18 20 41.2 32.6 59.5 54
2.3 5.1 18.8 20 26 32.9 62.7 55
2 5.6 46 20 30.5 32.9 45 55
2.7 6.3 21 20 49.9 33 46.4 56.2
3.5 6.6 26 20 19.6 33.2 53.2 56.7
3.9 7.6 28.8 20 43 33.6 60.8 57.1
6.7 8 21.3 20.2 44.3 33.7 68.1 57.4
8.6 8.4 5.3 20.2 27 33.9 68 58
13.2 8.5 18 20.6 34.5 34.9 62.3 58.2
12.6 9 14 20.7 46 35 45.3 60
10.7 9 27 20.8 30.6 35 48.1 60
21 9.9 10.1 20.8 49 35.5 60 60
8.7 10.1 14.1 21.1 42 35.6 69.9 60.1
7.7 10.3 12.7 21.2 28 36 72.6 61
4.2 10.4 10.6 21.3 51.4 36.1 63.6 61
18 10.5 12.5 21.5 37.5 36.2 53.3 61.6
30.8 10.8 12.5 21.5 37.5 36.2 88.7 62.6
8.2 10.8 18.3 21.9 62.3 36.7 72.3 63.3
4.6 10.8 20.7 22 47.6 36.7 54.7 65
9 11.2 9.5 22.1 34 37 70.3 65.2
26 11.4 24.6 22.4 63 38 59.1 66.2
10.9 11.4 13 22.8 43 38 64.5 67
32 11.9 24 23 29 38.4 53.8 67.9
3.8 12 24 23 32 39.3 50.8 68.3
9.1 12.7 11.8 23.1 51.2 39.4 104 69
8.2 12.7 9.7 24.2 33.4 39.4 70.7 70
9 13 33 24.7 33.4 39.4 72.8 70.8
17 13.2 28.3 25 47 40 60.9 74
10.2 13.5 26.6 25.4 40.7 40 103 74.6
9.7 13.7 43 25.5 43 40.8 68.1 75.7
6.4 13.8 27 25.9 49.1 42 68.1 75.7
20 14.2 27.3 26 42 42.2 69.8 76
9.7 14.2 28.5 26 46.1 42.4 92.7 76
14 14.5 28.8 26.4 46 42.6 72.4 76
6.5 14.6 42.6 26.5 37 43 51.2 78
9 14.7 21.3 26.6 49.9 43.5 105.4 80.7
5.4 15.3 21.3 26.6 49.9 43.5 80.3 85.7
10.8 15.9 28 27 38.3 44.5 82 86
17 16 20.3 27 53 44.7 76.8 88.7
12 16 18 27.1 46 45.2 66.1 90
10.3 16.1 33.5 27.3 68.9 47.8 93.4 91
17.5 16.2 28 27.3 49.2 48 70.6 91.1
12.5 16.3 37.5 27.5 42.7 48.2 85.6 99
6.4 16.3 28 27.6 48.1 48.5 97 101
19.5 16.5 25 28.1 44.2 48.7 129 103.3
9.3 16.7 32.6 28.6 50.4 48.9 88.1 104.9
21.7 17 27.1 28.6 38.9 49 107.9 117.7
17 17.1 33 29 49.7 50 144 122.5
12.4 17.3 24.5 29 54.2 50.3 134.5 159.9
18.6 17.3 35.9 29 46.9 50.4 178 160
6.8 17.3 62.1 29.4 64 50.6 153.8 164
24 17.8 33.4 29.9 51 51 93.2 166
6.6 17.8 50 30 45.3 51 202.7 171.3
20 18 24 30.3 36 51 139.3 174
10.4 18 36.1 31 37.5 51.1 159 174.9
33 18.1 38 31.1 59 51.2 178 211.3
9.9 18.2 20.4 31.2 69 51.2
28 18.3 22.2 31.3 51.2 51.9

References Aydin, A., Basu, A., 2005. The Schmidt hammer in rock material characterization. Engi-
neering Geology 81 (1), 1–14.
Adachi, J.I., Detournay, E., Drescher, A., 1996. Determination of Rock Strength Parame- Balci, C., Demircin, M.A., Copur, H., Tuncdemir, H., 2004. Estimation of optimum specific
ters from Cutting Tests. Proceedings of the 2nd North American Rock Mechanics energy based on rock properties for assessment of roadheader performance. Jour-
Symposium (NARMS 1996), pp. 1517–1523. nal of the South African Institute of Mining and Metallurgy 104 (11), 633–641.
Almenara, J.R., Detournay, E., 1992. Cutting Experiments in Sandstones with Blunt PDC Bieniawski, Z.T., 1968. The effect of specimen size on compressive strength of coal. Inter-
Cutters. In: Hudson, J.A. (Ed.), ISRM Symposium: Eurock, 92, pp. 215–220. Thomas national Journal of Rock Mechanics and Mining Sciences 5 (4), 325–335.
Telford, London, United Kingdom. Bieniawski, Z.T., 1974. Estimating the strength of rock materials. Journal of the South
ASTM, 2010. ASTM D7012-10 Standard Test Method for Compressive Strength and Elastic African Institute of Mining and Metallurgy 74 (8), 312–320.
Moduli of Intact Rock Core Specimens under Varying States of Stress and Temperatures. Bieniawski, Z.T., 1975. The point load test in geotechnical practice. Engineering Geology 9, 1–11.
 Author's personal copy

100 T. Richard et al. / Engineering Geology 147-148 (2012) 91–100

Broch, E., Franklin, J.A., 1972. Point-load strength test. International Journal of Rock Me- Huang, H., Lecampion, B., Detournay, E., 2012. Discrete element modeling of tool-rock
chanics and Mining Sciences 9 (6), 669–697. interaction I: Rock cutting. International Journal for Numerical and Analytical
Chaput, E.J., 1992. Observations and Analysis of Hard Rocks Cutting Failure Mechanisms Methods in Geomechanics. http://dx.doi.org/10.1002/nag.2113.
using PDC Cutters. Technical report, Imperial College, London, United Kingdom. Hucka, V., 1965. A rapid method of determining the strength of rocks in situ. Interna-
Chau, K.T., Wong, R.H.C., 1996. Uniaxial compressive strength and point load strength tional Journal of Rock Mechanics and Mining Science and Geomechanics Abstracts
of rocks. International Journal of Rock Mechanics and Mining Sciences and 2 (2), 127–130.
Geomechanics Abstracts 33 (2), 183–188. Hudson, J.A., Crouch, S.L., Fairhurst, C., 1972. Soft, stiff and servo-controlled testing ma-
Coudyzer, C., Richard, T., 2005. Influence of the Back and Side Rake Angles in Rock Cut- chines: a review with reference to rock failure. Engineering Geology 6 (3),
ting. AADE National Technical Conference and Exhibition, pp. 1–12. number AADE- 155–189.
05-NTCE-75, American Association of Drilling Engineers, Houston, Texas, U.S.A. Jaeger, J., Cook, N.G., Zimmerman, R., 1976. Fundamentals of rock mechanics. Chapman
Dagrain, F., 2006. Etude des Mecanismes de Coupe des Roches avec Couteaux Uses - and Hall.
Approche des mécanismes de frottement sous les couteaux par le concept du Kahraman, S., 2001. Evaluation of simple methods for assessing the uniaxial compres-
troisième corps. Ph. D. Thesis, Faculté Polytechnique de Mons. sive strength of rock. International Journal of Rock Mechanics and Mining Sciences
Dagrain, F., Richard, T., 2006. On the influence of PDC wear and rock type on friction co- 38 (7), 981–994.
efficient and cutting efficiency. Eurock 2006: Multiphysics Coupling and Long Term Nishimatsu, Y., 1972. The mechanics of rock cutting. International Journal of Rock
Behaviour in Rock Mechanics, pp. 577–584. Liege, Belgium, Taylor & Francis, London. Mechanics and Mining Sciences 9 (2), 261–270.
Dagrain, F., Poyol, E., Richard, T., 2004. Strength Logging of Geomaterials from Scratch Pamplona, M., Kocher, M., Snethlage, R., Barros, L.A., 2007. Drilling resistance: overview
Tests. EUROCK 2004 and 53rd Geomechanics Colloquium. Salsbourg, Austria Verlag and outlook. Zeitschrift Der Deutschen Gesellschaft Fur Geowissenschaften 158
Glückauf GmbH, Essen. (3), 665–676.
Deliac, E.P., 1986. Optimisation des Machines d'Abbatage à Pics. Ph. D. Thesis, Pells, P.J.N., 1993. Comprehensive Rock Engineering, vol. 3. Pergamon Press, pp. 67–85.
Universite Pierre et Marie Curie, Paris, France. chapter Uniaxial Strength Testing.
Detournay, E., Defourny, P., 1992. A phenomenological model for the drilling action of Perrier, R., 1997. Mesure de la durete des roches par une plaquette de diamant
drag bits. International Journal of Rock Mechanics and Mining Sciences and polycristallin. Mines et Carrieres 79, 32–38.
Geomechanics Abstracts 29 (1), 13–23. Reddish, D.J., Yassar, E., 1996. A new portable rock strength index test based on specific
Detournay, E., Drescher, A., Fourmaintraux, D., Defourny, P., 1995. Assessment of Rock energy of drilling. International Journal of Rock Mechanic Mining Science and
Strength Properties from Cutting Tests: Preliminary Experimental Evidence. Proceed- Geomechanic Abstracts 33 (5), 543–548.
ings of the Colloquium Mundanum on Chalk and Shales, pp. 1–10. Free University of Richard, T., 1999. Determination of Rock Strength from Cutting Tests. Master of Science
Brussels, Brussels, Belgium. Thesis, University of Minnesota, Minneapolis, Minnesota, U.S.A.
Detournay, E., Drescher, A., and Hultman, D.A., 1997. Portable rock strength evaluation Richard, T., Detournay, E., Drescher, A., Nicodeme, P., Fourmaintraux, D., 1998. The
device. United States Patent 5670711. Scratch Test as a Means to Measure Strength of Sedimentary Rocks. SPE/ISRM
Dey, T., Halleck, P., 1981. Some aspects of size-effect in rock failure. Geophysical Re- Eurock 98. Society of Petroleum Engineers, Trondheim, Norway, pp. 1–8. number
search Letters 8 (7), 691–694. SPE 47196.
Duc, M.N.M., 1974. Contribution à l' étude de la taille des roches. Ph. D. Thesis, Universite Rusnak, J., Mark, C., 1999. Using the Point Load Test to Determine the Uniaxial Com-
de Paris VI, Paris, France. pressive Strength of Coal Measure Rock. Proc. 19th Int. Conference on Ground Con-
Fairhurst, C., Lacabanne, W.D., 1957. Hard rock drilling techniques. Mine and Quarry trol in Mining, pp. 362–371.
Engineering 23, 157–161. Stavropoulou, M., 2006. Modeling of small-diameter rotary drilling tests on marbles.
Farmer, I.W., 1992. Main Lecture: Rock Testing — Deficiencies and Selection. In: International Journal of Rock Mechanics and Mining Sciences 43 (7), 1034–1051.
Rakowski, Z. (Ed.), Proceedings of the International Conference on Geomechanics Szwedzicki, T., 1998. Indentation hardness testing of rock. International Journal of Rock
1991. Balkema, Hradec, Ostrava, Czecho-Slovakia, pp. 3–7. Mechanics and Mining Sciences 35 (6), 825–829.
Fener, M., Kahraman, S., Bilgil, A., Gunaydin, O., 2005. A comparative evaluation of in- Thuro, K., Plinninger, R.J., Zah, S., Schutz, S., 2001. Scale effects in rock strength properties.
direct methods to estimate the compressive strength of rocks. Rock Mechanics Part 1: Unconfined compressive test and Brazilian test. In: Sarkka, P., Eloranta, P.
and Rock Engineering 38 (4), 329–343. (Eds.), ISRM Regional Symposium Eurock 2001. Balkema, Espoo, Finland, pp. 169–174.
Glowka, D.A., 1989a. The thermal response of rock to friction in the drag cutting pro- Tiano, P., 2001. The use of microdrilling techniques for the characterization of stone
cess. Journal of Structural Geology 11 (7), 919–931. materials. RILEM TC 177-MDT Workshop On-Site Control and Non-Destructive
Glowka, D.A., 1989b. Use of single-cutter data in the analysis of PDC bit designs: Part 1 - Evaluation of Masonry Structures, pp. 203–212. Mantova, Italy, RILEM Publications.
Development of a PDC cutting force model. Journal of Petroleum Technology 41 Tsiambaos, G., Sabatakakis, N., 2004. Considerations on strength of intact sedimentary
(8), 797–799 (844–849). rocks. Engineering Geology 72 (3–4), 261–273.
Hawkes, I., Mellor, M., 1970. Uniaxial testing in rock mechanics laboratories. Engineer- Tsur-Lavie, Y., Denekamp, S.A., 1982. Comparison of size effect for different types of
ing Geology 4 (3), 177–285. strength tests. Rock Mechanics 15 (4), 243–254.
Hoek, E., Brown, E.T., 1980. Underground Excavations in Rock. Institution of Mining and Ulusay, R., Hudson, J.A. (Eds.), 2007. The Complete ISRM Suggested Methods for Rock
Metallurgy, London, United Kingdom. Characterization, Testing and Monitoring:1974–2006. International Society for
Huang, H., Detournay, E., 2008. Intrinsic length scales in tool-rock interaction. ASCE In- Rock Mechanics.
ternational Journal of Geomechanics 8 (1), 39–44. Wawersik, W.R., Fairhurst, C., 1970. A study of brittle rock fracture in laboratory com-
Huang, H.Y., Detournay, E., Bellier, B., 1999. Discrete element modelling of rock cutting. pression experiments. International Journal of Rock Mechanics and Mining Science
In: Amadei, B., Kranz, R.L., Scott, G.A., Smeallie, P.H. (Eds.), The 37th U.S. Sympo- and Geomechanics Abstracts 7 (5), 561–564.
sium on Rock Mechanics (USRMS). Balkema, pp. 123–130.