Overview about Rockmechanical Lab

Testing – Part I: Mechanical Tests

Author: Dr. Thomas Frühwirt & Prof. Dr. habil. Heinz Konietzky (TU Bergakademie Frei-
berg, Geotechnical Institute)

1 Introduction.......................................................................................................... 1
2 Basic tests ........................................................................................................... 2
3 Mechanical lab tests ............................................................................................ 3
Static Hardness determination ...................................................................... 3
Rebound Hardness determination ................................................................. 5
Abrasion determination.................................................................................. 5
Point load index test ...................................................................................... 1
Sound velocity tests....................................................................................... 4
Uniaxial compression tests ............................................................................ 5
Uniaxial tension test ...................................................................................... 9
Brazilian test ................................................................................................ 12
Triaxial test (Karman-type) .......................................................................... 15
True triaxial tests ...................................................................................... 19
Shear box tests ........................................................................................ 20
Fracture toughness tests .......................................................................... 23
3.12.1 KI fracture toughness tests ................................................................ 23
3.12.2 KII fracture toughness tests ................................................................ 27
4 Acoustic emission monitoring ............................................................................ 28
5 Further dynamic testing methods ...................................................................... 30
6 Large-scale testing (physical models) ............................................................... 32
7 Literature ........................................................................................................... 34

Editor: Prof. Dr.-Ing. habil. Heinz Konietzky Layout: Angela Griebsch

TU Bergakademie Freiberg, Institut für Geotechnik, Gustav-Zeuner-Straße 1, 09599 Freiberg • sekr_fm@ifgt.tu-freiberg.de
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

1 Introduction
Rockmechanical lab testing provides basic parameters in respect to mechanical, hydrau-
lic and thermal rock properties. This also includes the interaction of these properties, ob-
tained by so-called hydro-thermal-mechanical coupled testing.
Several tests are similar to those common in soil mechanics [e.g. Germaine & Germaine
2009], but due to the high stiffness and strength level of the rock material as well as the
brittle characteristics of most of the rocks forces, resolution in respect to deformation and
displacement as well as reaction time of testing devices have to meet special require-
ments.

Rock mechanical lab testing can be subdivided into several types according to different
criteria, for instance:

 Static vs. dynamic tests

 Uniaxial vs. triaxial tests
 Coupled vs. uncoupled tests
 Sophisticated tests vs. indicator tests
 Standardized tests vs. scientific tests
 Compression tests vs. tensile tests
Standards for rockmechanical lab tests are formulated in regulations or recommendations
by national or international scientific organizations. Important documents are the follow-
ing:

 Recommendation of the DGGT (German Geotechnical Society)

 Recommendations of the ISRM (Int. Society for Rock Mechanics)
 ASTM standard (American Society for Testing and Materials)
Because brittle rocks have similar parameters and show similar pattern during testing
compared with concrete, standards and recommendation valid for concrete can be con-
sulted in addition to the rockmechanical literature.

Prerequisit to obtain reliable data from rockmechanical testing is the correct sample se-
lection and preparation. Samples should be prepared according to the specific recom-
mendations valid for the different tests. Important are size (e.g. length to height ratio or
ratio of grain size to sample size) and the surface conditions (e.g. roughness or parallel-
ism).
The most important mechanical properties are related to strength (e.g. uniaxial compres-
sive and tensile strength) and stiffness (e.g. Young’s or Deformation Modulus and Pois-
son’s ratio). The most important hydraulic parameters are porosity and permeability. The
most important thermal parameters are thermal conductivity and thermal expansion co-
efficient.

The next chapters describe the most popular tests in more detail. The authors are aware,
that several other test procedures exist. This chapter can cover only a few and the most
popular ones.

Seite 1 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

2 Basic tests
Before any more comprehensive test is performed, in most cases some basic parameters
have to be determined, like density, porosity and water content using simple test proce-
dures. Density values are obtained by determining the weight and the volume of the sam-
ple. One can distinguish between:

 Bulk density ρ
 Saturated density ρw
 Dry density ρd
 Grain density ρs
The grains (solid components of the sample) are characterized by mass Ms and volume
Vs. Pore water and corresponding mass are characterized by Mw and Vw. Pore volume is
given by Vv.

ρ=M/V
ρs = Ms / Vs
ρw = (Ms + ρwVv) / V
ρd = Ms / V

Based on these parameters also water content w, degree of saturation Sr, porosity n and
void ratio e, respectively, are given.

w = Mw / Ms
Sr = Vw / Vv
n = Vv / V
e = Vv / Vs
e = n / (1-n)

To determine dry density, samples have to be dried in an oven at 105°C for at least 24
hours, but at least until equilibrium in mass is obtained. The volume (rock matrix and
pores) can be determined by different methods: caliper method (volume determination by
means of mechanical measuring devices, e.g. caliper), mercury displacement method
(measurement of amount of mercury penetrating the dry sample), water displacement
method (measurement of amount of water penetrating the dry sample), buoyancy method
(difference in weight between saturated and dry sample).

More detailed information is given for instance in Ulusay & Hudson [2007].

Seite 2 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

3 Mechanical lab tests

Static Hardness determination

Static hardness testing is based on the intendation of a hard tool into the sample. Hard
metal balls or diamond cones or pyramides can act as indentors. The test load is applied
with a defined initial application time and duration and has to be applied perpendicular to
the sample. The intendation is measured after removal of the load, either with integrated
or separate optical devices (microscopes). The most popular methods applied in rock-
mechanics are the testing procedures according to Brinell (ball), Vickers (pyramide),
Knoop (pyramide) and Rockwell (different shapes). Sample thickness should be at least
10 times penetration depth of indentor. Load level and indentor size depend on material
and considered resolution. Hardness testing is standardized by national and international
regulations (e.g. ISO, DIN, ASTM). According to the different procedures different dimen-
sionless hardness parameters can be determined.

Fig. 3.1.1: Typical indentec hardness testing devices [company leaflets]

Seite 3 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.1.2: Multi-functional hardness testing device with digital image analysis [RML 2016]

Seite 4 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Rebound Hardness determination

The Schmidt Hammer is the most popular portable device to measure the rebound hard-
ness and is often used to estimate uniaxial compressive strength or Young’s modulus via
empirical relations [Ulusay 2015]. The Schmidt hammer consists of a spring loaded pis-
ton, which is released when the plunger is pressed against the surface. The percentage
of maximum stretched length of the spring before the release of the piston to the length
after rebound gives the so-called rebound hardness value R. Depending on rock strength
two different types are used: L-type (0.735 Nm) and N-type (2.207 Nm).

Fig. 3.2.1: Schmidt Hammer after (above) and before (below) releasing the spring [RML 2016]

Abrasion determination

Abrasion tests measure the resistance of steel pins to wear during defined interaction
with rock. During the past decades several testing procedures were developed to specify
the abrasion characteristics of rocks [e.g. Ulusay 2007]. The most popular procedure is
the so-called Cherchar test. During the test a steel pin (rockwell hardness of 54 – 56 and
tensile strength of 2 GPa) with angle of 60° and loaded by 70 N is scratching the rock
surface. A distance of at least 50 mm or even better 100 mm is recommended to deter-
mine the CAI (Cherchar Abrasivness Index) according to Plinninger et al. [2003]. CAI
itself is determined by microscopic inspection of the abrasion of the steel needle. CAI
varies between 0.3 (not very abrasive) and 6.0 (extremely abrasive). By performing sev-
eral scratch tests and rotating the sample anisotropy in abrasion can be determined.

Seite 5 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.3.1: Left side: Typical test device for determination of CAI.
Right Side: Scratching traces on limestone specimen after testing perpendicular to bedding
and in parallel to bedding planes, respectively [RML 2016]

Point load index test

The Point Load Index Test (PLT) is a simple alternative to the uniaxial compression test,
but does not deliver directly uniaxial or tensile compressive strength data, but rather an
index about rock strength, which can be either correlated to more precise parameters or
directly used in empirical design procedures. During PLT a rock sample is compressed
between conical steel platens (pins) until failure occurs. The PLT index IS is calculated
according to the following formulae:

IS = FS / A

where: FS is the failure load (force)

A is the corresponding area of the fracture plane (sometimes also the
square of the equivalent sample radius is used)

The PLT device (Fig. 3.4.1) should meet several requirements in respect to size, shape
and stiffness and allow to monitor the applied pressure including the determination of
peak pressure. Also, only those tests, which show unique fracture pattern should be used

Seite 1 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

for evaluation (Fig. 3.4.2). Because PLT index shows a significant scale effect (increase
with increasing sample size), a size correction has to be performed by log-log-plots or the
LOGAR-procedure [Thuro 2008]. If enough reliable data for one specific rock type exist,
a correlation between PLT index and uniaxial compressive strength (UCS) can be estab-
lished in terms of a linear relation: UCS = c*IS, where c is a correlation factor. Figure 3.4.3
and 3.4.4 show such correlations for Cottaer Sandstone and Innsbrucker Quarzphyllit,
respectively.
More detailed information about PLT testing are given in Ulusay & Hudson [2007] or
Thuro [2008].

Fig. 3.4.1: Point Load Index Device (loading frame, pump with manometer and data aquisition unit) [RML
2016]

Seite 2 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.4.2: Valid and invalid fracture pattern [Thuro 2008]

70.0
Data Set #1
Data Set #2
60.0 Data Set #3
Data Set #4

50.0
Uniaxial Compressive Strength [N/mm²]

σu = 43.2 ± 7.6 [N/mm²]

40.0

30.0

20.0

Is = 2.8 ± 0.5 [N/mm²]

10.0
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Point Load Strength [N/mm²]

Fig. 3.4.3: Correlation between Point Load Index and UCS, Postaer Sandstone [RML 2016]

Seite 3 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.4.4: Correlation between Point Load Index and UCS, Innsbrucker Quarzphyllit [Thuro 2008]

Sound velocity tests

Sound velocity is closely related to dynamic elastic constants by the following formulaes:

𝐸𝐸 (1 − 𝜈𝜈)
𝑉𝑉𝑝𝑝 = �
𝜌𝜌 (1 + 𝜈𝜈)(1 − 2𝜈𝜈)

𝐸𝐸 (1 − 𝜈𝜈)
𝑉𝑉𝑠𝑠 = �
𝜌𝜌 2(1 + 𝜈𝜈)
where:
Vp Compressional wave speed
Vs Transversal wave speed
E Dynamic Young’s modulus
ν Dynamic Possion’s ratio
ρ Density

Sound velocity is usally measured by piezo-electric sensors (transmitter and receiver) on

prepared cores samples, bars or bigger rock blocks. Frequencies are typically in the order
of several 100 kHz. Two different measuring procedures are common:

Seite 4 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

 Determination of wave speed of longitudinal and transversal wave

 Determination of velocity of dilatational and torsional waves in bar of rode-like
specimen
 Determination of resonant frequency of dilatational and torsional waves in bar of
rode-like specimen
Velocity of P- and S- waves are determined by dividing distance by travel time or multi-
plying 2 times the length of the bar with resonance frequency.

Despite the determination of dynamic elastic constants (E, ν) sound velocity measure-
ments can also be used to estimate the damage state, the anisotropy ratio or to classify
rocks.

Fig. 3.5.1: Typical equipment to measure sound velocity on core samples [RML 2016]

Uniaxial compression tests

The aim of uniaxial compression testing is to determine compressive strength and stiff-
ness of intact rock samples.

A suitable testing machine has to be used, which should have the following features (see
also specifications according to ASTM, DIN or ISRM / DGGT recommendations):
 Sufficient stiffness and load capacity
 Spherical seat and sufficient hardness of loading platens
 Sufficient size of loading platens (diameter of loading platen > sample diameter)
 High accuracy in measuring axial load and vertical displacement incl. display and
storage of data (measurement error < 1 %)

Seite 5 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

The rock samples should meet several requirements:

 Cylindrical shape with height to diameter ration of about 2 to 3
 Diameter of specimen should be related to the largest grain diameter by a ratio of
10:1 or bigger
 The ends of the specimen should be flat and not depart from perpendicularity
(parallel with high precision)
The testing shall be performed according to the following rules:
 Constant stress rate of about 0.5 to 1.0 MPa/s until failure
 To determine any kind of deformation modulus, Young’s modulus and Poisson’s
ratio, it is recommended to use separate high-precision measuring devices, like
LVDT’s, strain gauges or optical devices (accuracy in strain determination should
be better than 2 %)
Test reporting should contain information about the following items:
 Source of sample incl. lithological description and basic parameters (density, water
content etc.)
 Orientation of sample axis in respect to planes of anisotropy
 Sample geometry
 Date of testing, used machine and loading rate
 Failure pattern
 Measured (forces and displacements) and deduced values (peak strength, defor-
mations modulus etc.)
 Foto documentation of sample before and after testing
Uniaxial compressive strength is determined by dividing the peak load P by the initial
cross sectional area A0:

UCS = P / A0

Axial strain εa is measured by dividing change in axial length Δl by original axial length l0:

εa = Δl / l0

The lateral deformation can be given either as diametric strain εa (change in diameter) or
circumferential strain εc (change in circumference):

εd = Δd / d0 or εc = Δc / c0 (both give identical results)

Deformation moduli incl. Young’s modulus are given by the slope of the axial strain – axial
deformation curve, whereby according to the definitions different regions (parts) of the
stress – strain curve are used. Poisson’s ratio is given by dividing the horizontal strain εh
by vertical strain εa:

ν = εh / εa

Due to the inhomogeneity of the samples and measurement errors in general, it is rec-
ommended to perform at least 3 or better 5 tests per rock type. Scale (size) effects should
be taken into account (see also Fig. 3.6.5)

Seite 6 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.6.1: Typical machine for uniaxial compression tests [RML 2016]

Fig. 3.6.2: Sample, prepared for uniaxial compression testing with longitudinal and lateral strain
measurement direct on the sample [RML 2016]

Seite 7 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.6.3: Typical recording during uniaxial compression testing: vertical stress vs. vertical strain [RML
2016]

Fig. 3.6.4: Typical fracture pattern: axial splitting (left) and shear fracturing (right) [RML 2016]

Seite 8 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.6.5: Example for scale effect of sample size [Pinto, 1990]

Uniaxial tension test

For the uniaxial tension test, the same general rules and recommendations are valid as
for the uniaxial compression test (see chapter 3.6). After sample preparation cylindrical
metal caps shall be cemented to the specimen ends, so that tensile stress can be applied.
Care should be taken by cementation of the end caps and the load transfer system, so
that alignment of the whole system (sample + end caps) with the load axis is guaranteed
and any torsion or bending is avoided.

Seite 9 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.7.1: Uniaxial tension test: machine with sample during testing [RML 2016]

Seite 10 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.7.2: Samples with typical failure pattern [RML 2016]

Seite 11 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.7.3: Typical results obtained for Postaer Sandstone [Baumgarten & Konietzky, 2012]

Brazilian test

The Brazilian test (also called tensile splitting test) is an indirect method to determine the
tensile strength. This method is very popular, because sample preparation is easy and
standard compressive testing equipment can be used. During a Brazilian test a circular
rock disc is compressed by two diametrically applied forces (Fig. 3.8.1).

Seite 12 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.8.1: Stress situation during Brazilian test [Dinh 2011]

The applied compressive load should be transmitted to the sample only through a very
small arc of contact (< 10 degree). Thickness if disc should be half the diameter. Based
on a simplified analytical solution the splitting tensile strength σt can be calculated ac-
cording to the following formulae:

σt = 2P/(πDt)

where: P = maximum (peak) load at failure

D = diameter of disc
t = thickness of disc

The above mentioned formulae is only valid, if a central tensile crack is observed. For
anisotropic material, especially if plane of anisotropy is inclined to the loading direction,
mixed-mode or even predominant shear failure can occur (see Fig. 3.8.2). Also, some-
times shear failure is observed at the load entry points under the loading jaws. In these
cases the above mentioned formulae is not valid.

Fig. 3.8.2: Typical potential failure pattern for anisotropic material [Dinh & Konietzky 2014]

Seite 13 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.8.3: Typical test arrangement with arc-formed loading jaws and cylindrical sample [RML 2016]

Fig. 3.8.4: Typical fracture pattern during Brazilian tests for quasi-isotropic rocks (Postaer Sandstone) [RML
2016]

More detailed information, especially under consideration of anisotropic rocks and influ-
ence of testing parameters, is given by Dinh [2011], Dinh et al. [2013], Vervoort et al.
[2014], Dinh & Konietzky [2014].

Seite 14 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Triaxial test (Karman-type)

The Karman-type of testing is the most popular type of triaxial testing. This type of testing
is designed for cylindrical specimen and characterized by circumferential and axial pres-
sure (σ1 > σ2 = σ3 for compressional testing and σ1 < σ2 = σ3 for extensional testing).
Requirements on sample prepation are similar to those already mentiond in Chaper 3.6
(uniaxial compression testing). The axial pressure is applied by loading plates, the cir-
cumferential load by oil pressure. This demands, that the sample is encapsulated by a
flexible rubber sleeve to avoid any direct contact between the pressurized oil and the rock
sample. Sample size depends on triaxial cell size, but should follow a height to diameter
ratio of about 2. Figure 3.9.1 shows a sample prepared for triaxial testing including axial
and circumferential strain measurement sensors. Triaxial testing can be performed in
quite different ways, following different stress paths, deformation or stress controlled and
using different loading velocities. Also, due to the existing servo-algorithm, the post-peak
behavior (strain softening) can be observed, if the machine frame provides sufficient stiff-
ness (> 1 MN/mm). Fig. 3.9.3 illustrates the different stress paths, which can be applied
(CTC: conventional triaxial compression, RTC: reduced triaxial compression, RTE: re-
duced triaxial extension, HC: hydrostatic compression, TE: triaxial extension, RTE: re-
duced triaxial extension). To determine the complete failure envelope different proce-
dures can be applied: several single step tests, multi-stage tests or continuously failure
state tests (Fig. 3.9.4 and 3.9.5). National and international recommendations (e.g. given
by ISRM or DGGT) describe in detail the different procedures and demands for conduct-
ing triaxial tests. A comprehensive overview is given by Kwasniewski [2012].

Fig. 3.9.1: Sample with rubber sleeve prepared for testing with sensors to measure axial and
circumferential deformation [RML 2016]

Seite 15 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 2.9.2: Typical triaxial testing device with loading frame and triaxial cell, pressure unit and data
recording unit [RML 2016]

Fig. 2.9.3: Typical stress paths applied in Karman-type triaxial testing

Seite 16 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 2.9.3: Coventional single-stage test (axial pressure σ1 versus axial deformation ε1 or confining
pressure σ3)

Fig. 3.9.4: Multi-stage test (axial pressure σ1 versus axial deformation ε1 or confining pressure σ3)

Seite 17 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.9.5: Triaxial single step and continuous failure state tests for determination of failure envelope for
Postaer Sandstone, Germany (axial pressure σ1 versus axial deformation ε1 or confining pressure
σ3) [Baumgarten & Konietzky, 2012]

Seite 18 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 2.9.6: Change of compressional wave speed during triaxial testing of Postaer Sandstone, Germany
[Baumgarten & Konietzky 2012]

True triaxial tests

True triaxial testing allows the application of 3 different principal stresses (σ1 > σ2 > σ3)
on cubical samples. Depending on the specific construction 3 to 6 hydraulic cylinders or
pressure cells are necessary to apply the 3-dimensional stress state. True triaxial testing
is necessary to investigate the effect of the intermediate principal stress on strength, fail-
ure pattern and deformation characteristics. More detailed infos are given by Kwasniew-
ski [2012].

Seite 19 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Shear box tests

A shear box device consists of a loading frame, an upper and lower shear box and two
pistons to apply shear force and normal force (Fig. 3.11.1). Figure 3.11.2 shows a photo
of the central part of a sophisticated shear box device including the empty upper and
lower shear box. Inside these boxes the rock sample has to be placed. The fixation of the
sample (either cuboidal or cylindrical) is performed with special grout of high strength and
stiffness.

Vertical frame

Upper shear frame

Piston for normal load

150

Piston for shear load

Lower shear frame

Fig. 3.11.1: Principal sketch of a shear box device [Konietzky et al. 2012]

Fig. 3.11.2: Foto of a shear box device and view into empty shear boxes [Konietzky et al. 2012]

Seite 20 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Shear box testing can be performed in quite different manner:

 as constant normal load (CNL) test or constant normal stiffness (CNS) test (Fig.
3.11.3)
 as static, cyclic or dynamic test
 as pure mechanical test or HM-coupled test
 as test for intact rock samples or test of discontinuities (joints, fractures etc.)
 as single-stage or multi-stage test
 as test up to the peak strength or until the residual strength
During the test the following parameters should be monitored:
 axial and horizontal forces
 axial displacement and heave of loading plate (dilation measurement)
In addition surface roughness of shear plane can be scanned before and after the tests
as well as during test breaks. Normal stress and shear stress can be determined by the
recorded forces and the corresponding areas, whereby it should be considered, that ef-
fective shear area may reduces with ongoing shear displacement and therefore, the nor-
mal stress should be up-dated continuously.

N = constant
(a) (b)
K

S S

Fig. 3.11.3: Principle of CNL (left) and CNS (right) testing

Classical evaluation of shear tests include the determination of cohesion, friction and di-
lation. Cohesion and friction are determined using the Mohr-Coulomb theory by linear
regression over several data pairs of normal and shear stresses (Fig. 3.11.4 and
Fig. 3.11.5).

Seite 21 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

9
8 3.6 MPa
7 5.0 MPa
Shear stress [MPa]

10.0 MPa
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10
Shear displacement [mm]
Fig. 3.11.4: Example: Shear test results for 3 tests with different normal stress of 3.6, 5.0 and 10.0 MPa

9
8
y = 0.568x + 2.4616
7
Shear stress [MPa]

2
R = 0.9129
6
5
Peak
4
Residual y = 0.5106x + 0.1676
3
R2 = 0.9959
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11
Normal stress [MPa]
Fig. 3.11.5: Example: Determination of peak and residual strength for a slate by linear regression of 3
shear tests (peak cohesion = 2.5 MPa, peak friction angle 29°, residual cohesion = 0.16 MPa,
residual friction = 27°)

The dilation Ψ is defined as ratio between measured vertical to horizontal displacement

components (valid under the assumption, that the shear plane is horizontal):

Ψ = Δun / Δus

However, one should take into consideration the actual orientation of the fracture plane.
If fracture plane is inclined, ‘apparent’ dilation is observed and true value should be ob-
tained by correction (Fig. 3.11.6).

Seite 22 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Z
σno ∆un < 0 σno

Slope direction Slope direction

Slope direction
Shear
θ (positive) direction

Shear
direction
O Initial position ∆us X

Z σno
σno ∆un > 0

Slope direction
Slope direction θ (negative) Shear
direction
Slope direction
Shear
direction
O Initial position ∆us X
Fig. 3.11.6: Problem of ‘apparent’ dilation due to inclined fracture plane

Fracture toughness tests

Three different types of fracture toughness can be distinguished:

 Mode-I: tensile fracture
 Mode-II: in-plane shear fracture
 Mode-III: out-of-plane shear fracture (torsion fracture)
So far, for practical applications only Mode-I and Mode-II fracture propagations are con-
sidered. Therefore, lab testing concentrates on these two types:
 Determination of critical stess intensity factor (= fracture toughness) KIC
 Determination of critical stess intensity factor (= fracture toughness) KIIC
In material sciences several methods were developed to determine fracture toughness.
Due to the specific material characteristics of rocks specific testing methods were devel-
oped in rockmechanics.

3.12.1 KI fracture toughness tests

Most popular methods are:
 Chevron Bend Specimen (CBS) test
 Short Rod Specimen (SRS) test
 Cracked Chevron Notched Brazilian Disc (CCNBD) test
 Semi-circular bend (SCB) test
CBS test is similar to a 3-point-bending test, but needs a special prepared cylindrical
sample. A special shaped notch [Ulusay 2007] has to be created at the center of the

Seite 23 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

specimen opposite to the load entry point. The measurements can be conducted on two
different levels:
 Level-1-testing: considers only load at failure and sample dimensions
 Level-2-testing: considers non-linearities during fracture propagation based on
additional measurement of crack opening displacement and corresponding cor-
rection terms in calculating fracture toughness [Ulusay 2007]
KIC based on level-1-testing is determined by the following formulae:

𝐴𝐴𝐴𝐴
𝐾𝐾𝐼𝐼𝐼𝐼 [𝑀𝑀𝑀𝑀𝑀𝑀√𝑚𝑚] =
𝐷𝐷
𝑎𝑎 𝑎𝑎 2 𝑆𝑆
𝐴𝐴 = [1.835 + 7.15 + 9,85 � � ]
𝐷𝐷 𝐷𝐷 𝐷𝐷

where:
F maximum load (load at failure) in kN
D diameter of specimen in cm
S distance between support points (3.33 * D) in cm
A Chevron tip distance from specimen surface (0.15*D) in cm

Fig. 3.12.1.1: Test set-up for CBS test [RML 2016]

Seite 24 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

SRS tests are designed for using short rods (cylindrical core specimen of short length)
with chevron-shaped notch cut along core axis. SRS test can also be performed as level-
1 or level-2 test. KIC based on level-1-testing is determined by the following formulae:
𝐹𝐹
𝐾𝐾𝐼𝐼𝐼𝐼 [𝑀𝑀𝑀𝑀𝑀𝑀√𝑚𝑚] = 24 1.5
𝐷𝐷
where:
F maximum load (load at failure) in kN
D diameter of specimen in cm

CCNBD testing is similar to the classical Brazilian test, but needs a special preparation
of the sample: a special designed notch (Fig. 3.12.1.2). KIC is determined by the following
formulae:
𝑃𝑃√𝑎𝑎
𝐾𝐾𝐼𝐼𝐼𝐼 = 𝑁𝑁𝐼𝐼
√𝜋𝜋𝜋𝜋𝜋𝜋
where:
R radius of specimen
a half crack length
P load at failure
B thickness of specimen
NI special function depending on sample dimension and crack orientation α (see
Chen & Konietzky [2014])

SCB tests is a special three-point bending test as shown in Fig. 3.12.1.5. More specific
recommendations are given by Kuruppu et al. (2014).

𝑃𝑃√𝜋𝜋𝜋𝜋
𝐾𝐾𝐼𝐼𝐼𝐼 = 𝑌𝑌
2𝑟𝑟𝑟𝑟 𝐼𝐼
where:
P maximum load
a notch length
r radius of sample
t thickness of specimen
YI special dimensionless function depending on sample dimension and crack orien-
tation α according to Kuruppu et al. [2014]

Seite 25 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.12.1.2: CCNBD sample configuration [Aliha & Ayatollahi 2014]

Fig. 3.12.1.3: Test set-up for CCNBD test [RML 2016]

Fig. 3.12.1.4: Test set-up for SCB test [Kataoka & Obara 2015]

Seite 26 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

3.12.2 KII fracture toughness tests

Several test methods have been developed to measure KIIC [e.g. Backers 2004, Backers
& Stephansson 2012, Chen & Konietzky 2014, 2015, Ulusay 2015). Some commom
methods used in rockmechanics are illustrated in Fig. 3.12.2.1. The Central Cracked
Notch Brazilian Disc (CCNBD) test is quite interesting, because easy to perform using
the standard equipment for Brazilian tests. Also, the same test arrangement can be used
to determine both, Mode-I and Model-II fracture toughness (Chen & Konietzky 2014,
2015).

Fig. 3.12.2.1: Mode-II fracture toughness methods: A: antisymmetric four point bending, B: antisymnmetric
four point bending cube, C: punch through shear, D: compression shear cube, E: short beam compression,
F: centrally cracked Brazilian disc, G: triaxial compression, H: three point bending semi disc [Backers 2004]

Seite 27 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

4 Acoustic emission monitoring

Crack development and fracture propagation are connected with radiation of seismic en-
ergy (seimic waves). Therefore, acoustic sensors can be used to monitor damage pro-
cesses. Typical AE sensors have the following characteristics:

 frequency range between 103 to 106 Hz

 velocity proportional registration (sometimes also acceleration proportional)
 sensors based on piezo-ceramic material
 sensor size: a few mm to a few cm

Sensors are either directly connected to the sample or fixed at the loading frame or laod-
ing plates. Monitoring can be performed in quite different ways, but in principal all applied
techniques including data processing and evaluation are similar to earthquake monitor-
ing. With one or just a very few sensors only event counting including some relative mag-
nitude evaluation can be performed. If a complete network is installed, localization can
be performed and seismic source parameters can be determined, like seismic moment,
magnitude, fault plane solution, stress drop, source dislocation, source dimensions. Lo-
calization can be performed with different techniques using first arrivals of P- and S-
waves. Sophisticated AE analysis is described by Stanchits et al. (2011, 2014).

Typical AE systems consists of the following elements:

 Sensors
 Pre-amplifier
 Transient recorder
 Computer with software (on-line data evaluation and postprocessing)

To perform localization and seismic source parameter determination a well spherical ar-
ranged network of sensors (at least 6 to 10) is necessary. Also, noise level should be as
low as possible, because only events above the noise level can be detected.

Seite 28 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 4.1: Typical AE workplace with transient recorder and on-line display [RML 2016]

Fig. 4.2: AE localization at different stress level for uniaxial loading [Liu 2015].

Seite 29 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 4.3: Fracture toughness test with AE monitoring und localization [RML 2016]

5 Further dynamic testing methods

Several rockmechanical applications or processes demand the consideration of dynamic
parameters, like earthquake engineering, blasting, explosions, rockbursts, drilling, impact
and collision problems etc. Dynamic rock properties deviate significantly from static ones
as exemplary demonstrated in Fig. 5.1. Dynamic rock parameters have to be always
given together with the loading rate used during testing, because they are strongly de-
pending on loading rate. Most popular dynamic testing methods are [Xia & Yao 2015,
Zhou et al. 2012]:

 Split Hopkinson pressure bar (SHPB) test

 Dynamic notched semi-circular bend (NSCB) test
 Dynamic Brazilian test
 Dynamic compression test

Main dynamic rock parametes are:

 Dynamic compressive strength

 Dynamic tensile strength
 Dybnamic punch shear strength
 Dynamic bending strength
 Dynamic fracture toughness
 Dynamic Young’s modulus
 Dynamic Poisso’s ratio

Seite 30 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 5.1: Tensile and punch shear strength for sandstones as a function of loading rate [Xia & Yao 2015]

Many of the dynamic tests are performed in the same way as the corresponding static
tests, but on much higher loading rate. However, the very popular Split Hopkinson pres-
sure bar test requires a special experimental set-up as exemplary shown in Figure 5.2.
SHPB consists of three interacting bars: a striker bar, an incident bar and a transmitted
bar. Via a special gun the striker bar is accelerated. The impact of the striker bar on the
free surface of the incident bar induces longitudinal compressive waves in two directions.
The left propagation wave is fully released at the striker bar and forms the trailing edge
of the incident compressive pulse. Reaching the incident bar – sample interface, part of
the wave is reflected and the remainder passes through the specimen. Strain gauges are
used to measure the stress wave pulse. The objective of SHPB tests is to determine the
dynamic stress strain curves and to deduce dynamic strength and stiffness of the mate-
rial.

Fig. 5.2: Pricipal sketch of Split Hopkison pressure bar device [Xia et al. 2015]

Seite 31 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

6 Large-scale testing (physical models)

The aim of large-scale experiments is to investigate scale effects, bigger rock blocks or
the behavior of more complex systems. Large scale experiments consist of large loading
frames. Two types of tests can be distinguished:

 Physical models based on up-scaling of complex systems under consideration of

the laws of physical equivalence (e.g. mining systems or dam constructions)
 Test of geotechnical elements in real size (e.g. railroad systems, support ele-
ments or huge rock blocks)
In most cases these test are unique and need a special designed measuring and loading
arrangement. Large scale testing is very time and cost consuming and therefore often
replaced or at least supported by numerical simulations.

Fig. 3.1: Loading frame with physical model of tunnel with installed anchors (State key lab, Zhengzhou,
China)

Seite 32 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Fig. 3.2: Static and dynamic loading rig for testing of railway foundation in real size (CSU, Changsha,
China)

Fig. 3.3: True triaxial cell for large samples up to 3 x 3 x 3.5 m (State key lab, Zhengzhou, China)

Seite 33 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

7 Literature
Aliha, M.R.M.; Ayatollahi, MN.R. (2014): Rock fracture toughness study using cracked
chevron notched Brazilian disc specimen under pure modes I and II loading – A stochas-
tical approach, Theoretical and Applied Fracture Mechanics, 69: 17-25.
Backers, T. (2004): Fracture toughness determination and micromechanics of rock under
mode I and mode II loading, PhD thesis, University Potsdam, Germany
Backers, T.; Stephansson, O. (2012): ISRM suggested method for the determination of
model II fracture toughness, Rock Mech Rock Eng, 45: 1011-1022
Baumgarten, L.; Konietzky, H. (2012): Stress-strain, strength and failure behavior of
Postaer Sandstone in tension and compression tests – laboratory investigations and nu-
merical modelling with PFC3D, Veröffentl. Institut Geotechnik (ed. H. Konietzky), TU Berg-
akademie Freiberg, Heft 2012-1, p. 41-61
Chen, W., Konietzky, H. (2014): Simulation of heterogeneity, creep, damage and life time
for loaded brittle rocks, Tectonophysics, DOI 10.1016/j.tecto.2014.06.033
Chen, W., Konietzky, H. (2015): Simulation of time-independent and time-dependent frac-
turing in sandstone, Engineering Geology, 193: 118-131
Dinh, Q.D. (2011): Brazilian test on anisotropic rocks – laboratory experiment, numerical
simulation and interpretation, Veröffentl. Institut Geotechnik (ed. H. Konietzky), TU
Bergakademie Freiberg, Heft 2011-2
Dinh, Q.D; Konietzky, H.; Herbst, M., (2013): Brazilian Tensile Strength Tests on some
Anisotropic Rocks, Int. J. Rock Mech. Mining Sci., 58: 1-7.
Dinh, Q.D.; Konietzky, H. (2014): Numerical simulations and interpretations of Brazilian
tensile tests on transversely isotropic rocks, Int. J. Rock Mech. Min. Sci., 71: 53-63
Germaine, J.T & Germaine, A.V. (2009): Geotechnical laboratory measurements for en-
gineers, John Wiley & Sons, 351 p.
Kataoka, M. & Obara, Y. (2015): Anisotropy in fracture toughness od sedimentary and
crystalline rocks estimated by semi-circular bend test, Proc. Eurock-2015
Konietzky, H.; Frühwirt, T.; Luge, H., (2012): A new large dynamic rockmechanical direct
shear box device, Rock Mechanics Rock Engineering, 45(3): 427-432.
Kuruppu et al. (2014): ISRM-suggested method for determining the mode I static fracture
toughness using semi-circular bend specimen, Rock Mech. Rock Eng., 47: 267-274
Kwasniewski, M. (2012): Recent advances in studies on the strength of rocks under gen-
eral triaxial compression conditions, Veröffentl. Institut Geotechnik (ed. H. Konietzky), TU
Bergakademie Freiberg, Heft 2012-2, p. 77-104
Liu, J. et al. (2015): Moment tensor analysis of acoustic emission for cracking mecha-
nisms in rock with a pre-cut circular hole under uniaxial compression, Eng. Fracture Me-
chanics, 135: 206-218
Mittelbach, L.; Konietzky, H.; Baumgarten, L., (2012): Ultrasonic wave measurements
during triaxial tests - laboratory tests and numerical simulations, BAW-Mitteilungen, 95:
71-78.

Seite 34 von 40
 Overview about Rockmechanical Lab Testing
Only for private and internal use! Updated: 15 January 2016

Nguyen, V.M. (2013): Static and dynamic behavior of joints in schistose rock: Lab testing
and numerical simulation, Veröffentl. Institut Geotechnik (ed. H. Konietzky), TU
Bergakademie Freiberg, Heft 2013-3
Nguyen, V.M; Konietzky, H.; Frühwirt, T., (2014): New Methodology to Characterize
Shear Behavior of Joints by Combination of Direct Shear Box Testing and Numerical
Simulations. Geotech Geol Eng.
Pinto da Cunha, A., (1990): Scale Effects in Rock Masses. International Workshop 7. &
8.6.1990, Leon, Norway, Balkema
Ulusay, R. & Hudson, J.A. (2007): The complete ISRM suggested methods for rock char-
acterization, testing and monitoring: 1974-2006, Commission on Testing Methods, Int.
Society for Rock Mechanics
Ulusay, R. (2015): The ISRM suggested methods for rock characterization, testing and
monitoring: 2007-2014, Springer, DOI 10.1007/978-3-319-07713-0
Tan, X.; Konietzky, H.; Frühwirt, T., (2014): Laboratory observation and numerical simu-
lation of permeability evolution during progressive failure of brittle rocks., Int. J. Rock
Mech. Mining Sci., 68: 167-176.
Plinninger, R.; Käsling, H.; Thuro K.; Spaun, B. (2003): Testing conditions and geome-
chanical properties influencing the Cherchar abrasiveness index (CAI), Int. J. Rock Mech.
Mining Sci., 40: 259-263.
RML (2015): Rock Mechanical Laboratory, Chair for Rock Mechanics, Geotechnical In-
stitute, TU Bergakademie Freiberg, Germany

Stanchits, S. et al. (2011): fracturing of porous rock induced by fluid injection, Tectono-
physics, 503: 129-145
Stanchits, S. et al. (2014): Onset of hydraulic fracture initiation monitored by acoustic
emuission and volumetric deformation measurements, Rock Mech. Rock Eng., 47: 1521-
1532
Thuro, K. (2008): The new sugested Method No. 5 of the AK 3.3. – Point Load Index
Tests on Rocks Samples, Veröffentl. Institut Geotechnik (ed. H. Konietzky), TU Bergaka-
demie Freiberg, Heft 2008-3, p. 23-36
Vervoort, A.; K.-B., Min; Konietzky, H.; Cho, J.-W; Debecker, B.; Dinh, Q. D.; Frühwirt, T.;
Tavallali, A., (2014): Failure of transverselly isotropic rock under Brazilian test conditions.,
Int. J. Rock Mech. Mining Sci., 70: 343-352
Xia,K; Yao, W (2015): Dynamic rock test using split Hopkinson (Kolsky) bar system – A
review, Journal of Rock Mech. Geoetechn. Eng., 7: 27-59
Zhou, YX. et al. (2012): Suggested methods for determination the dynamic strength pa-
rameters and mode-I fracture toughness of rock material, Int. J. Rock Mech. Min. Sci.,
49: 105-112

Mode detailed infos and technical parameters of rockmechanical testing equip-

ment can be found under:
http://tu-freiberg.de/fakult3/gt/feme/labor.en.html

Seite 35 von 40