GSI From Different Methods
GSI From Different Methods
GSI From Different Methods
DOI 10.1007/s10706-015-9880-x
ORIGINAL PAPER
Received: 14 November 2014 / Accepted: 4 April 2015 / Published online: 11 April 2015
Springer International Publishing Switzerland 2015
Abstract The main purpose of this paper is to based on the Q and RMi inputs is especially sensitive
present the results of a comparative analysis of the GSI to the infilling and roughness of discontinuities, while
values predicted by different empirical equations in the methods based on the RMR1989 ratings the
currently available in literature which apply the input discontinuity characteristics influencing the GSI value
parameters used in the best known rock mass classi- the most demonstrate to be the aperture, the roughness
fication systems, namely the RMR1989, the Q and the and the length. The analyses have mainly highlighted
RMi. For this aim, probabilistic analyses based on the possible uncertainties still related to the quantita-
Monte Carlo simulations have been developed using, tive estimation of the GSI and allowed for the
as inputs, the geomechanical field data collected for a recognition of the input rock mass parameters that
real rock mass. Afterwards, the GSI outputs obtained may have the highest impact on the GSI in the
from the different empirical approaches have been different estimation methods.
statistically analyzed and compared. The results of
simulations indicate that the diverse relationships may Keywords GSI Quantitative methods Monte
predict dissimilar values of the GSI for the same rock Carlo simulations Sensitivity analysis
mass. In general, the highest values have been
obtained from the equations which apply the
RMR1989 input ratings, while the methods based on
RMi produced the lowest results. Sensitivity analyses 1 Introduction
performed on the simulation outputs show that, for the
examined case study, the input parameters reflecting The Geological Strength Index (GSI) represents today
the degree of jointing of the rock mass, namely the the most worldwide used engineering index for
RQD and the rock block volume, have the largest categorizing the rock mass quality for input into
effects on the calculated GSI, while those describing continuum numerical analysis codes and closed form
the characteristics of discontinuities show lesser solutions based on the HoekBrown failure criterion
influence and may depend on the equation adopted. (Marinos and Hoek 2000; Marinos et al. 2007; Carter
In particular, the GSI estimated from the methods 2010).
The index was introduced in the mid-nineties as a
qualitative approach to relate the HoekBrown crite-
G. L. Morelli (&)
rion to geological field observations in hard rocks
Geotecna Progetti S.r.l., Milan, Italy
e-mail: [email protected]; (Marinos et al. 2007; Day et al. 2012) and has been
[email protected] successively extended to accommodate also the most
123
984 Geotech Geol Eng (2015) 33:983995
variable of rock masses, including extremely poor 1989), the Q (Barton et al. 1974) and the RMi
quality sheared rock masses of weak schistose mate- (Palmstrom 1995).
rials and the typical heterogeneous lithological for- Quantitative methods can be used successfully for
mations such as flysch and molasses (Hoek et al. 1998; estimating GSI in rock masses that can be though of as
Marinos and Hoek 2000, 2001; Hoek et al. 2005; blocky, which behaviour mainly depends on sliding
Marinos et al. 2006). and rotation of intact rock pieces and is controlled by
The GSI system was specifically formulated to the condition of discontinuities that separate indi-
attempt to characterize rock masses from a more vidual blocks. On the other hand, such methods
geological perspective, so as to better meet the need become inappropriate for rocks lying at both ends of
for delivering reliable input data related to those the rock competence scale and for tectonically
properties considered to govern the mechanical be- disturbed rock masses in which the structural fabric
haviour of a rock mass (Carter and Marinos 2014). has been destroyed, for which the original descriptive
According to Marinos and Hoek (2000) and Mari- techniques are recognized to work best (see Marinos
nos et al. (2007), the heart of the GSI system is a et al. 2007; Carter et al. 2008; Hoek et al. 2013; Carter
careful engineering geology description of the rock and Marinos 2014).
mass which is currently based first upon the lithology Independently from the estimation method adopted,
and secondary on structure (or blockiness) and con- the GSI may show high spatial variability in a rock
dition of the discontinuity surfaces assessed from mass mainly in response to the inherent variability of
visual examination of rock exposures. the degree of jointing and of the discontinuities
As such, the estimation of the GSI number, which characteristics and, thus, it can be regarded as a
theoretically varies in the nominal range 0100, is random variable following a particular probability
essentially a qualitative procedure whose accuracy distribution (e.g. Cai 2011).
may, therefore, depend on the level of experience and Conceptually, one of the best ways to investigate the
judgment of the observer directly involved in field possible variation of a random variable is to utilize
mapping. probabilistic simulations, by which the variability of
Carter (2010) also pointed out as a main problem the input parameters used to calculate the output
with rock mass classification misuse in civil contract parameter can be consistently modelled and quantified.
works may arise from classification mismatch due to In rock engineering, examples in which strength and
subjective observational differences which, in some deformability of rock masses have been evaluated
cases, can result from the manipulation of classifica- through a probabilistic approach have become, year on
tion parameter selection made in order to maintain year, more and more numerous (e.g. Kim and Gao 1995;
consistency of perceived geological description suit- Hoek 1998, 1999; Russo et al. 2001; Potsch et al. 2004;
able for the observers contractual perspective. Sari 2009; Sari et al. 2010; Cai 2011; Idris et al. 2013).
For such reasons, the issue of quantifying the GSI However, none of these works specifically focused on
has become, over the years, more and more a priority the comparison between the GSI values theoretically
and several attempts to reduce subjectivity in GSI obtainable from different quantitative methods. Actu-
estimation and to benchmark its range started to emerge ally, a first attempt has recently been made by Sari
in literature. In particular, quantitative methods for (2014), although limited to the comparison between few
quantify GSI based on a numerical quantification of the of the available equations.
parameters defining the rock mass structure in terms of Taking inspiration from this issue, the main purpose
blockiness and interlocking degree and of the condition of this paper is to provide a more comprehensive
of discontinuities, have been fine-tuned and presented investigation of the possible variability of the GSI
by some Authors (see e.g. Hoek et al. 1995; Sonmez index calculated using the empirical relationships
and Ulusay 1999; Cai et al. 2004; Cai and Kaiser 2006; most commonly applied in the current rock engineer-
Russo 2007, 2009; Day et al. 2012; Hoek et al. 2013). ing practice.
All these methods are generally based on empirical For this goal, probabilistic analyses based on the
equations which apply input parameters and ratings Monte Carlo simulation methodology have been
borrowed from the best known rock mass classification performed using, as inputs, consistent geomechanical
index systems, principally the RMR (Bieniawski field data collected for a real rock mass and the GSI
123
Geotech Geol Eng (2015) 33:983995 985
output datasets obtained by applying the different properties (Barton et al. 1974; Bieniawski 1989;
equations statistically analyzed and compared. Grimstad and Barton 1993; Barton 2002, 2006; Galera
et al. 2007; Barton and Bieniawski 2008). Further
adaptations and modifications to the original classifi-
2 Empirical Relationships in Use for Estimating cations also occurred over the years, mainly aimed at
the GSI mining, surface slopes, dam design and TBM applica-
tions (see e.g. Laubscher 1977; Romana 1985; Sapigni
A number of empirical relationships to quantitatively et al. 2002; Romana 2003; Ozdemir and Nilsen 1993;
estimate the GSI have been progressively proposed in Barton1999; Bieniawski et al. 2007, 2008).
literature since the appearance of the GSI index in the The RMi index was principally developed to
middle of the nineties. characterize the strength of a rock mass for use in
Those well-known and more commonly applied rock engineering and design. Its value is also applied
today have been considered in the present work and as input for estimating stability and rock support in
are summarized in Table 1. underground openings and for determining the con-
As can be discerned, the methods at hand can be stants of the HoekBrown failure criterion for rock
divided in three main groups: one including the masses (see Palmstrom 1995, 1996a, b, 2000; Palm-
relationships based on the rock mass input parameters strom and Singh 2001).
used in the RMR1989 classification system (Bieniawski All classification systems are essentially based on
1989), another encompassing those using the parameters ratings for the parameters reflecting rock mass jointing
applied in the Q-system (Barton et al. 1974), and the last and discontinuity conditions.
including the expressions employing the input pa- Although there are similarities between these sys-
rameters of the RMi index system (Palmstrom 1995). tems, actually they present differences in the way the
As well known, the RMR and Q systems originated, input parameters are combined to calculate the refer-
and have been specifically updated mainly for estimat- ence index (i.e. the RMR index derives from an addition
ing tunnel support in civil engineering applications and of ratings while Q and RMi indexes use a combination
only later they were extended for assessing rock mass of multiplication, division and exponential calculation)
123
986 Geotech Geol Eng (2015) 33:983995
and in the input parameters included, in addition to their discontinuities, specifically the length, the aperture, the
scores and weights in calculation. roughness, the infilling and the weathering.
As a consequence, correlations between classifica- The RMR1989 index theoretically range from a
tion indexes have been proved to present large minimum of 0 to a maximum of 100 and is, then,
inaccuracies (Palmstrom 2009) and the necessity of consistent with the GSI scale.
comparing different independent systems for charac-
terizing a rock mass has been recurrently recommend- 2.2 Q-System
ed by many Authors as the most suitable approach
during the design process (see e.g. Carter 1992; The Q-system (Barton et al. 1974) was originally
Palmstrom and Broch 2006; Barton and Bieniawski developed as an empirical design method for estimat-
2008; Palmstrom 2009; Carter 2010; Carter and ing the rock supports in tunnels and caverns using
Marinos 2014). many underground excavation case histories. The
A short compendium of the principal characteristics value of the Q index used to classify the rock mass is
of the three classification systems is reported in the defined by six parameters combined in the following
following. equation
RQD Jr Jw
2.1 RMR System Q 2
Jn Ja SRF
The Rock Mass Rating system (1989 version) (Bieni- where RQD = given as the value for the pa-
awski 1989) provides a classification of the rock mass rameter;Jn = rating for the number of joint
in five classes of geomechanical quality on the base of sets;Jr = rating for the joint roughness and wavi-
the RMR1989 index. This index is calculated as the sum ness;Ja = rating for the joint alteration and infill-
of six basic input ratings (R1R6) related to selected ing;Jw = rating for ground water;SRF = rating for
rock mass parameters: the UCS of intact rock material the rock mass stress situation.The Q index theoretical-
(R1), the RQD (R2), the spacing (R3) and condition ly spans in a very wide range, indicatively from 0.001
(R4) of the most relevant discontinuities, the ground to 1000 or more.
water conditions (R5) and the orientation of the main
discontinuity set with respect to the engineering work 2.3 RMi System
under design (R6).
The classification index is obtained as The RMi index (Palmstrom 1995) expresses an
approximate measure of the uniaxial compressive
RMR1989 R1 R2 R3 R4 R5 R6 1 strength of a rock mass, given as
The basic ratings for calculating the RMR1989 are RMi rc JP 3
schematized in Table 2.
The rating R4 specifically reflects the condition of the where rc is the uniaxial compressive strength of the
rock mass discontinuities and is calculated as the sum of intact rock material (in MPa) measured on samples
five sub-ratings (R4ae) each related to a property of with diameter 50 mm and JP is the Jointing
123
Geotech Geol Eng (2015) 33:983995 987
Parameter, expressing the reduction in strength of the covariance between variables is recognized to play
intact rock caused by joints. The JP scores the main an important role in probabilistic analyses, the issue
joint features of a rock mass and is calculated as seems still not extensively dealt with in rock engineer-
ing literature and many authors assumed independence
JP 0:2 jC 0:5 VbD 4 of parameters due to lack of specific research results
with the coefficient D equal to (e.g. Hoek 1998; Russo 2009; Cai 2011; Idris et al.
2013; Sari 2009, 2014).
D 0:37 jC 0:2 5 An attempt to investigate the effect of incorporating
relationships between rock mass parameters in Monte
where jC is the joint condition factor, which
Carlo simulations and to check their influence on the
incorporate the characteristics of roughness/undula-
estimation of RMR index and HoekBrown pa-
tion, alteration/infilling and length of joints
rameters has recently been undertaken by Sari
jR jL (2012). In his work, a correlation matrix defining the
jC 6
jA relationships between inputs was constructed basing
on opinions and judgment of some distinguished
with jR = jS 9 jW = joint smoothness factor (small
experts in rock engineering and engineering geology.
scale roughness) 9 joint waviness factor (large scale
The main conclusion was that the consideration of
roughness). The factor jR is similar to the Jr of the Q
parameter correlations would not significantly affect
system;jL = joint length factor;jA = joint alteration
the results generated by Monte Carlo simulations, and
factor (similar to the Ja number in the Q system).Vb is
only a slight decrease of the standard deviations of the
the elementary rock block volume given in m3.Being
output rock mass parameters was observed.
an expression of the rock mass compressive strength
the RMi index hasnt range limitation.
3.2 Derivation of Rock Mass Input Parameters
123
988 Geotech Geol Eng (2015) 33:983995
RQD 110 2:5Jv 7 representative value of 60, typical for moderately long
and flat prismatic blocks (Palmstrom 2005) has been
with RQD = 0 for Jv [ 44 and RQD = 100 for assumed for b.
Jv \ 4. The frequency histograms and probability density
To translate the original RMR1989 ratings for the functions that best describe the distributions and
discontinuity characteristics, directly assessed during variability of field mapping data are summarized in
field surveys, to the corresponding ratings of the Q and Table 3 for each input parameter used to calculate the
RMi classification systems, the reference tables pur- GSI.
posely presented by Palmstrom (2009) and reproduced In the table, the parameters reflecting the discon-
in Appendix, have been adopted. tinuity characteristics specifically refer to the descrip-
The values for the unitary rock block volume (Vb) tive terms adopted in the RMR1989 classification
of the rock mass have been estimated using the system (Bieniawski 1989).
simplified relation of Palmstrom (1995) To synthesize the main characteristics of the
Vbm3 b Jv3 8 rock mass, the histograms in Fig. 1 display the
distributions for RQD and discontinuity spacing
where b is a block shape factor (Palmstrom 1995). obtained from the statistical elaboration of the survey
Basing on field observations and measurements, a data.
Table 3 Rock mass parameters used as inputs in Monte Carlo simulations for estimating the GSI
Parameter Distribution type (Mean SD) or frequency
UCS (MPa) Normal (100 35; truncated at min = 35 MPa and max = 200 MPa)
Number of joint sets 3 % three sets; 97 % four or more sets
RQD (%) Normal (82.22 7.41; truncated at max = 100 %)
Block volume Vb (cm3) Lognormal (65,928.12 61,632.02)
Joint spacing (m) Normal (0.34 0.10; truncated at min = 0.2)
Discontinuity length 20 % 13 m; 80 % 310 m
Discontinuity aperture 18 % none; 5 % \ 0.1 mm; 74 % 0.11 mm; 3 % 15 mm
Discontinuity roughness 11 % rough; 83 % slightly rough; 6 % smooth
Discontinuity infilling 95 % none; 5 % hard friction materials
Discontinuity weathering 2 % unweathered; 98 % moderately weathered
Fig. 1 Statistical distributions of field mapping data for RQD (left) ad joint spacing (right)
123
Geotech Geol Eng (2015) 33:983995 989
Fig. 2 Statistical distributions of data for JCond89 (left) and Peak Friction Angle (right) of discontinuities
4 Simulation Results
Table 4 Summary statistics of the GSI values resulting from Monte Carlo simulations
Equation (1) Equation (2) Equation (3) Equation (4) Equation (5) Equation (6) Equation (7)
GSI-RMR1989 GSI-RMR1989 GSI-Q (Hoek GSI-Q GSI-Q (Hoek GSI-RMi GSI-RMi
(Hoek et al. (Hoek et al. et al. 1995) (Barton et al. 2013) (Cai-Kaiser (Russo
1995) 2013) 1995) 2006) 2009)
123
990 Geotech Geol Eng (2015) 33:983995
123
Geotech Geol Eng (2015) 33:983995 991
b Fig. 4 Sensitivity charts for the input parameters used to Carter 1992; Palmstrom 1995, 1996; Carter and
calculate the GSI in the different quantitative methods Marinos 2014).
Also the rating for the uniaxial compressive
strength (UCS) of intact rock, when contemplated,
As can be observed, the results of analysis indicate shows a prominent influence on the GSI.
that the different empirical relationships may predict Among the parameters expressing the characteris-
dissimilar values for the same rock mass. tics of discontinuities, which show rather limited
In general, the relationships that use the RMR1989 ranges in the examined dataset, the most relevant in
input ratings (Eqs. 1 and 2) show the highest values for calculations seem to vary from one method to another
the GSI, while those based on RMi parameters (Eqs. 6 likely as a result of the different suite of parameters
and 7) produce the lowest results. considered in the diverse classifications and of the
In particular, the RMi-based equation of Cai and weight each parameter assumes in the relationships.
Kaiser (2006) (Eq. 6) provides the minimum mean In particular, for the rock mass examined it is
value of 48.9, which is about 29 % lower than the observed that when the equations based on the
maximum of 69.2 obtained from the RMR1989 expres- RMR1989 system are used, the aperture, the roughness
sion of Hoek et al. (2013) (Eq. 2). and the length demonstrate the greatest impact on the
The equations based on the Q-System mostly GSI, while the joint infilling and weathering show
predict intermediate mean values, with the exception lesser influence.
of the one proposed by Hoek et al. (2013) (Eq. 5) On the other hand, the infilling and the roughness
which demonstrate to fit better with the results of the parameters, in addition to the length of discontinuities
RMR1989 correlations. when considered, exhibit the highest influence in the
It is worth noting that the GSI calculated using the methods based on the Q and RMi systems.
relationship of Russo (2009) (Eq. 7) shows the highest In general, in all applied methods the weathering
Coefficient of Variation, thus indicating a tendency for parameter demonstrates a low impact.
such a method to provide more scattered results. On
the other hand, the traditional Q-based expressions
(Eqs. 3 and 4) demonstrate the lowest CVs. 6 Conclusions
All GSI datasets generated by the probabilistic
simulations seem to follow normal or lognormal The results of performed simulations have indicated
distributions. that the different empirical relationships currently in
use for calculating the GSI may predict dissimilar
5 Sensitivity Analysis values for the same rock mass.
In general, the highest values were derived from the
Sensitivity analyses have been performed to find out, expressions which apply the RMR1989 input ratings,
for the examined case study, the input parameters while the methods based on the RMi parameters
influencing the GSI value the most. invariably predicted the lowest results. The equations
For this purpose, the contribution to variance based on the Q-system mostly produced intermediate
method, which allows the sensitivity information to be mean values, with the exception of the one proposed
calculated and displayed as the percent of the output by Hoek et al. (2013) which demonstrated to fit better
variance contributed by each input parameter, has with the results of the RMR1989 correlations.
been used on simulations results. Divergences are principally linked to the inherent
The findings of analysis are presented in the differences existing between the different classifica-
sensitivity charts of Fig. 4. tion systems upon which the equations for calculating
It is apparent from the graphs that, irrespective of the GSI are based; these differences mainly relate to
the equation used, the parameters reflecting the degree the diverse origin and structure of the classification
of jointing of the rock mass, namely the RQD and the schemes and to the type of input parameters consid-
unitary block volume (Vb), are invariably the most ered, in addition to their weights in calculations (see
influential parameters in defining the GSI, rather in Carter 1992; Tzamos and Sofianos 2007; Palmstrom
accordance with some previous statements (see e.g. 2009).
123
992 Geotech Geol Eng (2015) 33:983995
Divergences also arise from the known inaccura- and descriptions. Hence, the accuracy of its evaluation
cies in correlating parameters used to quantify the in the field may be influenced by the sound experience
same rock mass property in the different classification and judgment of the observer. In this light, particular
systems, in particular those describing the degree of care must be taken to ensure that adequate attention is
jointing of the rock mass (i.e. RQD/Jn, RQD, joint given to defining this parameter during field mapping,
spacing and block volume) and the roughness, weath- for which the execution of specific tests to character-
ering and infilling of discontinuities, which are ize the physical and mineralogical properties of the
differently expressed and quoted in the diverse infilling material is highly recommended.
classification schemes (see e.g. Barton 1987; Carter The findings of the presented study highlight the
1992; Milne et al. 2002; Palmstrom 2005; Palmstrom possible uncertainties still related to the quantitative
and Broch 2006; Palmstrom 2009). estimation of the GSI and suggest that the application
For the examined case study, sensitivity analyses of more than one independent approach would be
showed that, irrespective of the equation adopted, the advisable to provide a more accurate estimation of this
GSI is most sensitive to the input parameters which index and a better definition of its possible variability.
reflect the rock mass jointing, namely the RQD and the However, in doing this it is essential to be well-
elementary block volume (Vb), therefore highlighting conscious of the inaccuracies possibly connected to
the need for a careful and reliable quantification of the use of the different quantitative estimation meth-
these parameters during field characterization. ods and to be clearly aware of the single input rock
As far as the joint characteristics are concerned, the mass parameters which may have, for each method,
GSI estimated from the methods based on the Q and the highest impact on the calculated GSI.
RMi inputs is sensitive to the ratings assigned to the Moreover, as the reliability of these quantitative
infilling and roughness of discontinuities, while in the methods mostly dependent on the correct evaluation of
methods based on the RMR1989 the discontinuity the input parameters, a careful engineering geological
parameters influencing the GSI value the most are the description of the rock mass carried out by experi-
aperture, the roughness and the length. enced and well-trained observers, who should also
In this regard it has to be outlined that the have a sound geological background, still remains an
parameter infilling of discontinuities, which plays indispensable prerequisite to reducing subjectivity
a primary role also in the quantification of the joint variation in GSI quantification.
alteration factors (Ja, jA) used in Q and RMi
systems, conversely to other measurable discontinuity
parameters like e.g. length, aperture and roughness Appendix
(e.g. Priest 2004; Palmstrom 2001; Morelli 2014), is
essentially assessed through qualitative observations See Table 5.
Table 5 Combined ratings for discontinuity characteristics used in the RMR, Q and RMi systems (from Palmstrom 2009)
Discontinuity characteristics RMR1989 Q RMi
123
Geotech Geol Eng (2015) 33:983995 993
Table 5 continued
Discontinuity characteristics RMR1989 Q RMi
123
994 Geotech Geol Eng (2015) 33:983995
123
Geotech Geol Eng (2015) 33:983995 995
Milne D, Palmstrom A, Peck W (2002) Reply to Bartons Romana M (1985) New adjustment ratings for application of
comments on GeoEng2000 workshop summary. Published Bieniawski classification to slopes. In: Proc. Int. Symp. On
in ISRM News 72, 2002. Under Letter to the Editor of the the Role of Rock Mech., 4953
ISRM News Journal. 18 February 2002 regarding Journal Romana M (2003) DMR, a new geomechanics classification for
article of December 2001 use in dams foundations, adapted from RMR.4th Interna-
Morelli GL (2014) On joint roughness: measurements and use in tional Symposium on Roller Compacted Concrete (RCC)
rock mass characterization. Geotech Geol Eng 32(2): Dams, Madrid
345362 Russo G (2007) Improving the reliability of GSI estimation: the
Ozdemir L, Nilsen B (1993) Hard Rock Tunnel Boring Pre- integrated GSI-RMi system. In: Proceedings I.S.R.M.
diction and Field Performance. Rapid Excavation and workshop Underground Works under Special Condi-
Tunneling Conference (RETC) Proceedings, Boston, USA tions, Madrid: 123130
Palmstrom A (1995) RMi A rock mass characterization system Russo G (2009) A new rational method for calculating the GSI.
for rock engineering purposes. PhD thesis, University of Tunn Undergr Space Technol 24:103111
Oslo, Norway, pp. 400. http://www.rockmass.net Russo G, Kalamaras GS, Origlia L, Grasso P (2001) A
Palmstrom A (1996a) Characterizing rock masses by the RMi probabilistic approach for characterizing the complex
for use in practical rock engineering. Part 1: the develop- geologic environment for design of the new metro do Porto.
ment of the rock mass index (RMi). Tunn Undergr Space AITES-ITA 2001 World Tunnel Congress Progress in
Tech 11(2):175188 Tunneling after 2000. Milano, June 1013, 2001 Volume
Palmstrom A (1996b) Characterizing rock masses by the RMi III: 463470
for use in practical rock engineering. Part 2: some practical Sapigni M, Bert M, Bethaz E, Busillo A, Cardone G (2002)
applications of the rock mass index (RMi). Tunn Undergr TBM performance estimation using rockmass classifica-
Space Tech 11(3):287303 tions. Rock Mech Min Sci 39:771788
Palmstrom A (2000) Recent developments in rock support es- Sari M (2009) The stochastic assessment of strength and de-
timates by the RMi. J Rock Mech Tunn Technol 6:19 formability characteristics for a pyroclastic rock mass. Int J
Palmstrom A (2001) Measurement and characterization of rock Rock Mech Min Sci 46:613626
mass jointing. In: Sharma VM, Saxena KR (eds) In-situ Sari M (2012) Stochastic estimation of the Hoek-Brown strength
characterization of rocks. A. A. Balkema Publishers, parameters using spreadsheet models. In: EUROCK 2012,
Brookfield, pp 4997 ISRM International Symposium, 2830 May 2012,
Palmstrom A (2005) Measurements of and correlations between Stockholm
block size and rock quality designation (RQD). Tunn Un- Sari M (2014) Incorporating variability and/or uncertainty of
dergr Space Technol 20:362377 rock mass properties into GSI and RMi systems using
Palmstrom A (2009) Combining the RMR, Q and RMi classi- Monte Carlo method. In: G. Lollino et al. (eds.), Engi-
fication systems. http://www.rockmass.net, pp. 25 neering geology for society and territory-volume 6: applied
Palmstrom A, Broch E (2006) Use and misuse of rock mass geology for major engineering projects, Part XIV, Paper
classification systems with particular reference to the 152. Springer International Publishing Switzerland 2015:
Q-system. Tunn Undergr Space Technol 21:575593 843849
Palmstrom A, Singh R (2001) The deformation modulus of rock Sari M, Karpuz C, Ayday C (2010) Estimating rock mass
masses: comparison between in situ tests and indirect es- properties using Monte Carlo simulation: ankara andesites.
timates. Tunn Undergr Space Technol 16(3):115131 Comput Geosci 36:959969
Potsch M, Schubert W, Goricki A, Steidl A (2004) Determina- Sonmez H, Ulusay R (1999) Modifications to the geological
tion of rock mass behaviour type: a case study. EUROCK strength index (GSI) and their applicability to the stability
2004, 53rd Geomechanics Colloquium. Schubert (ed.) of slopes. Int J Rock Mech Min Sci 36:743760
Priest SD (2004) Determination of discontinuity size distribu- Tzamos S, Sofianos AI (2007) A correlation of four rock mass
tions from scanline data. Rock Mech Rock Eng classification systems through their fabric indices. Int J
37(5):347368 Rock Mech Min Sci 44(4):477495
123