=.

J

22

•

Classification of Rock Masses for Engineering: The RMR System and Future Trends

Z. T. BIENIAWSKI

Pennsylvania State University, University Park, PA, USA

22.1 INTRODUCTION 553
22.2 PURPOSES OF ROCK MASS CLASSIFICA nON 554
22.3 CLASSIFICATION PARAMETERS 554
22.4 MAJOR CLASSIFICATIONS CURRENTLY IN USE 555
22.5 THE ROCK MASS RATING SYSTEM 555
22.5.1 Definition of the RM R System 556
22.5.2 Classification Procedure 556
• 22.5.3 Applications of the RM R System 565
22.5.3.1 Rock mass deformability 565
22.5.3.2 Rock mass strength 566
22.5.3.3 Advantages and limitations 568
22.5.3.4 An example of use 569
22.5.4 Database for the RMR System 569
22.5.5 Correlations with Other Classification Systems 570
22.6 FUTURE TRENDS 572
22.7 SUMMARY AND CONCLUSIONS 572
22.8 REFERENCES 572 •

22.1 INTRODUCTION r~;,~J...~

Rock mass classifications form the backbone of the empirical design approach which relates practical experience gained on previous projects to the conditions anticipated at a proposed site. They are widely employed in rock engineering [1]. In fact, on many projects, the classification approach serves as the only systematic and practical basis for the design of complex excavations in rock. Most of the tunnels, mines, slopes and foundations constructed currently make use of a classification system.

The best known classification system is the Terzaghi rock load classification, introduced over 45 years ago [2]. Since then this classification has been modified [3] and new rock mass classification systems proposed. These systems took cognizance of the new advances in rock support technology, namely rock bolts and shotcrete, as well as addressing different engineering projects such as tunnels, chambers, mines, slopes and foundations. Today there are many rock classification systems in existence and the major ones are listed in Table 1. These classifications have been applied throughout the world - in the USA [1-4], Canada [5], Europe [6-8J and elsewhere [9-14].

553

22.2 PURPOSES OF ROCK MASS CLASSIFICATION

•

Rock mass classifications were never intended as the ultimate solution to design problems, but only as a means towards this end. In fact, some 20 years ago when work started on the major classification systems now in use, the underground excavation scene worldwide was characterized by haphazard site investigation programs and by limited, if any, design procedures. Rock mass classifications were developed to create some order out of chaos in site investigation procedures and to provide desperately needed design aids. They were not intended to replace analytical considerations, field observations, or engineering judgment; they were simply to be design aids, forming only a part of the rock engineer's 'bag of tools'. It is therefore appropriate to identify clearly the purpose and benefits, as well as the pitfalls, of rock mass classifications before proceeding with a description of any specific classification system.

Rock masses are classified for the following purposes: (i) to identify the most significant parameters influencing the behavior of a rock mass; (ii) to divide a particular rock mass formation into a number of rock mass classes of varying quality; (iii) to provide a basis for understanding the characteristics of each rock mass class; (iv) to derive quantitative data for engineering design; (v) to recommend support guidelines for tunnels and mines; (vi) to provide a common basis for communication between engineers and geologists; and (vii) to relate the experience on rock conditions at one site to the conditions and experience encountered at others.

The above items suggest three main benefits of rock mass classifications: (i) improving the quality of site investigations by calling for the minimum input data as classification parameters; (ii) providing quantitative information for design purposes; and (iii) enabling better engineering judgment and more effective communication on a project.

When used correctly and for the purposes for which they were created, rock mass classifications can be powerful aids in rock engineering. When abused, they can be counterproductive. The major pitfalls are: (i) using rock mass classifications as the 'ultimate empirical cook book', i.e. ignoring analytical and observational design methods; (ii) using one rock mass classification system only, i.e. without cross-checking the results with at least one other system; (iii) using rock mass classifications without enough input data; and (iv) using rock mass classifications without full realization of their conservative nature and their limits arising from the database on which they were developed .

•

22.3 CLASSIFICATION PARAMETERS

•

The output from any rock mass classification is only as good as the input data. It is therefore important that the necessary basic input data are available and reliable.

Unlike other engineering materials, rock presents unique problems. First of all, rock is a complex material varying widely in its properties, and in most rock engineering situations not one but a number of rock types will be present. Furthermore, a choice of rock materials is only available if there is a choice of alternative sites for a given project, although it may be possible, to some extent, to reinforce the rock surrounding the excavation. Most of all, the rock engineer and geologist are confronted with rock as an assemblage of blocks of rock material separated by various types of discontinuities, such as joints, faults, bedding planes, etc. This assemblage constitutes a rock mass. Consequently, the engineering properties of both intact rock and the rock mass must be considered.

The question immediately arises as to how the rock material is related to the rock mass. In answering this question one must note, first of all, that the importance of the properties of intact rock material will be generally overshadowed by the properties of the discontinuities in the rock masses. However, this does not mean that the properties of the intact rock material should be disregarded when considering the behavior of jointed rock masses. After all, if discontinuities are widely spaced or if the intact rock is weak and altered, the properties of the intact rock may strongly influence the gross behavior of the rock mass. Furthermore, a sample of a rock material sometimes represents a small-scale model of the rock mass, since they have both gone through the same geological cycle. Nevertheless, in general, the properties of the discontinuities are of greater importance than the properties of the intact rock material.

The selection of the parameters of greatest significance for assuring rock mass stability is an important issue. There appears to be no single parameter which can fully and quantitatively describe a jointed rock mass for engineering purposes. Various parameters have different significance and only taken together can they describe a rock mass satisfactorily.

The strength of the rock material is considered as the first necessary parameter because it constitutes the strength limit of the rock mass. The uniaxial compressive strength of rock material

Classification of Rock Masses for Engineering: The RM R System and Future Trenas )))

can be determined in the field indirectly by means of index tests so that one is not restricted to laboratory testing.

The second parameter most commonly employed is the rock quality designation (RQD). This is a quantitative index [3J based on a modified core recovery procedure which incorporates only sound pieces of core 100 mm or greater in length. The RQD is a measure of drill core quality or fracture frequency, and disregards the influence of joint tightness, orientation, continuity and gouge (infilling). Consequently, the RQD does not fully describe a rock mass.

Other essential parameters used in describing rock masses are: spacing of discontinuities; condition of discontinuities (roughness, continuity, separation, joint-wall weathering, infilling); orientation of discontinuities; groundwater conditions (inflow, pressure); and in situ stresses.

It is accepted that in the case of surface excavations and those near-surface tunnels which are controlled by the structural geological features, the strength of the intact rock material, the spacing, condition and orientation of the discontinuities and the groundwater conditions are important parameters. In the case of mines or deep tunnels and chambers where the behavior of rock masses is stress controlled, knowledge of the virgin stress field or the changes in stress can be of greater significance than the geological discontinuities. Most civil engineering projects will fall into the first category of geologically controlled rock mass structures.

In essence, the following parameters are considered to be most important for rock mass classification [15]: (i) the strength of the rock material, which constitutes the upper strength limit ofthe rock mass. Because of the prominence of the compressive stress in rock engineering problems, the uniaxial compressive strength is the usual parameter; (ii) the rock quality designation (RQD), a quantitative index widely used in rock engineering, affording comparisons of rock behavior in varied engineering situations; (iii) the basic geological parameters usually included in any geological survey, such as spacing, orientation and condition (i.e. roughness, separation, continuity, weathering and infilling) of discontinuities; (iv) groundwater conditions; (v) stress field; and (vi) major faults and folds.

A thorough discussion of the methods for quantitative description of discontinuities in rock masses will be found in Chapters 8-11 of this reference work, and in a document issued by the ISRM [16].

•

22.4 MAJOR CLASSIFICATIONS CURRENTLY IN USE

Of the many rock mass classification systems in existence today, six should be mentioned because they are important contributions, namely those proposed by Terzaghi [2J, Lauffer [6J, Deere [3J, Wickham et al. [4J, Bieniawski [17J and Barton et al. [7].

The 1946 rock load classification of Terzaghi [2J was the first practical classification system introduced and has been dominant in the USA for many years, proving very successful for tunneling with steel supports. Lauffer's classification [6J was a considerable step forward in the art of tunneling in 1958 since it introduced the concept of the stand-up time of an unsupported span in a tunnel, which is highly relevant in determining the type and amount of tunnel support. Deere's classification [3J of 1967 introduced the rock quality designation (RQD) index - a simple and practical method of describing and classifying the quality of rock core recovered from boreholes.

The concept of rock structure rating (RSR), developed in the United States in 1972 by Wickham et al. [4J, was the first system featuring classification ratings for weighing the relative importance of classification parameters. The RMR system (also called the Geomechanics Classification), proposed by Bieniawski in 1973 [9J, and the Q system proposed by Barton et al. [7J in 1974, also include different parameter ratings but, in addition, both provide quantitative data for the selection of modern tunnel reinforcement measures such as rock bolts and shotcrete. Moreover, in addition to mining the RMR system has been applied to rock slopes and foundations and ground rippability assessments [ll

Today, the RMR system and the Q system are the two rock mass classifications most commonly used in rock engineering around the world .

•

22.5 THE ROCK MASS RATING SYSTEM

The Rock Mass Rating (RMR) system, otherwise known as the Geomechanics Classification, was developed by the author [9J during 1972-73. It was modified over the years as more case histories became available and to conform with international standards and procedures [17]. Over the past 20 years, the RMR system has stood the test of time and benefited from extensions and applications

by many authors throughout the world. These varied applications, amounting to 351 case histories [1], involved tunnels, chambers, mines, slopes and foundations. Nevertheless, it is important that the RMR system is used for the purpose for which it was developed, and not as a 'cookbook' for empirical design.

• 22.5.1 Definition of the RMR System

Due to the RMR system having been modified several times, and since the method is interchangeably known as the Geomechanics Classification or the Rock Mass Rating system, it is important to state that the system has remained essentially the same in principle in spite of the changes. Thus, any modifications and extensions were the outgrowth of the same basic method and should not be misconstrued as new systems. To avoid any confusion, Table 1 lists all the extensions of the RMR system which were valuable new applications but still part of the same overall RMR system [10, 14, 18-22].

Furthermore, some users of the RMR system list their results as 'CSIR rating' or talk of the 'CSIR Geomechanical' system. This is incorrect and it has never been used or suggested to this effect by the author. The correct expressions are 'Rock Mass Rating system', 'RMR system' or the 'Geomechanics Classification'. While it is true that the author has worked for an organization whose initials were CSIR, that group did not develop the system and most of the work on his system was performed after he left the CSIR over 15 years ago.

•

22.5.2 Classification Procedure

The following six parameters are used to classify a rock mass usmg the RMR system

(Geomechanics Classification).

1. Uniaxial compressive strength of rock material.

2. Rock quality designation (RQD).

3. Spacing of discontinuities.

4. Condition of discontinuities.

Table 1 Major Rock Mass Classifications Currently in Use [1]

N arne of classification

Originator Country of
and date origin
Terzaghi, 1946 USA
Lauffer, 1958 Austria
Rabcewicz, Pacher and Austria
Miiller, 1964
Deere, 1967 USA
Wickham et al.; 1972 USA
Bieniawski, 1973 South Africa
(last modified, 1979, USA)
Laubscher, 1977 South Africa
Ghose and Raju, 1981 India
Kendorski et al.; 1983 USA
Serafim and Pereira, 1983 Portugal
Gonzales de Vallejo, 1983 Spain
Una~ 1983 USA
Romana, 1985 Spain
Newman, 1985 USA
Venkateswarlu, 1986 India
Robertson, 1988 Canada
Barton et al., 1974 Norway
Franklin, 1975 Canada
International Society for Rock
Mechanics, 1981 Applications

Rock loads Stand-up time NATM

Tunnels with steel support Tunneling

Tunneling

Core logging, tunneling

Rock quality designation RSR concept RMR system

( Geomechanics Classification)

RMR system extensions

Tunneling

Tunnels, mines, slopes, foundations

Mining Coal mining

Hard rock mining Foundations Tunneling

Roof bolting/coal Slope stability Coal mining

Coal mining

Slope stability Tunnels, chambers Tunneling

General, communication

•

Q system Strength-Size Basic geotechnical

description

•

~:;

I

i:

•.. :

.

,

.,.

5. Groundwater conditions.

6. Orientation of discontinuities.

To apply the RMR system, the rock mass is divided into a number of structural regions such that certain features are more or less uniform within each region. Although rock masses are discontinuous in nature, they may, nevertheless, be uniform in regions when, for example, the type of rock or the discontinuity spacings are the same throughout that region. In most cases, the boundaries of structural regions will coincide with major geological features such as faults, dykes, shear zones, etc. After the structural regions have been identified, the classification parameters for each structural region are determined from measurements in the field and entered onto the input data sheet shown in Figure 1.

The RMR system is presented in Table 2. In section (a) of Table 2, five parameters are grouped into five ranges of values. Since the various parameters are not equally important for the overall classification of a rock mass, importance ratings are allocated to the different value ranges of the parameters, a higher rating indicating better rock mass conditions. The importance ratings are assigned to each parameter according to section (a) of Table 2. In this respect, the average typical conditions are evaluated for each discontinuity set and the ratings are interpolated, using the classification charts in Figures 2, 3, 4 and 5, and the guidelines in Table 3. The charts are helpful for borderline cases and also remove the impression that abrupt changes in ratings occur between categories. Figure 5 is used if either RQD or discontinuity data are lacking. Based on the correlation data from Priest and Hudson [37], Figure 5 enables an estimate of the missing parameter. Furthermore, it should be noted that the importance ratings given for discontinuity spacings apply to rock masses having three sets of discontinuities. Thus, when only two sets of discontinuities are present a conservative assessment is obtained. In this way, the number of discontinuity sets is considered indirectly. Laubscher [10] presented a rating concept for discontinuity spacings as a function of the number of joint sets. It can be shown that when less than three sets of discontinuities are present, the rating for discontinuity spacing may be increased by 30% .

Unioxiol compressive strength (MPo)

Figure 2 The RMR system: ratings for the strength of intact rock material

ROD ("Iol

Figure 3 The RMR system: ratings for RQD

•

•

~

w fCJ)

>CJ)

(!) Z f<l:: a:

CJ) CJ)

<l:: ~

:.:::

U o a:

z o f<l::

U u,

CJ) CJ)

::i

u CJ) u

z <l:: :r: u w

~ o w

o

~ a: o u,

~

-o c

f:::> a.. Z

(/) W

E

::J Z

~ Z o o (/)

o u, o z o

E

o z o o

E E E E E

y

(")000

1 ...... ('.1('.1

~ 6 1\

co c:: co

a.

o

0) c::

iii

:J <:l c:: :J

-c

Q) 0. 0.

~

'"

~ E E E E E co

E E E E E '"

Q)

<.J co 't :J

'"

(/) W

E

::J Z

i= z o o (/)

c

:.:: " u CTl ~o 'x CTl a: .~ o.-l C ~ -l :::l(/)«

~

~

« o -l « z o

E

o o « o z « (/) :.:: a: « ~

W a: -l « a:

W z W (!)

.~

CTl <.J

..2

£

<.J Q) 0.

'"

c:

Q)

E

:J <.J o <:l

~ a: (/)

~

~

:::l Z i= z o u

··· . 0 d a:

~

::J «

:::l o 1--+----1 W

a: o o -l

-l

a: o

W a..

~

::.::

U o a:

E :::t: ~ a.. W o

-l « a:z ::Jo ~U(!)

::JW a: a: ~

(/)

c o

to

c 0) .c:;;

<0 o

~

'c :0

:0 ". ~

::0

~

:~

. Ol· .

.9- ~ .9- .~ .9-

o~o 0 0

.~

CTl :J

o

.><: <.J o a:

II o d

a:

-l

-s

a:

W

~

~ :.:: u L..----4o

a: ~ o

~.

z

13

Q)

'0' a.

Q)

-5 '0

Q)

E

co z

.. >~..a ~<:l :J Q)

"'0

-:J

o <:l Gi CD c:: ... .!:'! 0 ro (/)uo

u, o :::t: ~ o z W a: ~ (/)

c::

.2 iii

c:: 0)

'iii Q)

o

(/) z o i= « ~ z W

a: o a.. o o z «

W :.::

a: ~ (/)

~

CD .0) .CTl

.Qi

.> 'CTl

.~

9

'" c

.2 o ~

<:l

(/) W

E

:::l Z i= z o u (/)

o u, o o z

U « a.. (/)

... CD (/)

E E E E E E

g g E ~ '"jI E

000 0<0<0 N Y

a:

W

~

o Z ::J o a: (!)

CTl a.. .><:

(/) W (/) (/) W a: ~ (/)

:::l !:::

(/)

z

•

0> C

o 0::

Spacing of discontinuities (rnrn)

Figure 4 The RMR system: ratings for discontinuity spacing

•

100 90 80 70

35

40

o o 0::

Numbers in graph

represent combined ROD and spacing ratings of each region

----Average correlation line

20 30 40 60 100 200 600

Mean discontinuity spacing (rnrn)

2000

Figure 5 The RMR system: correlation between RQD and discontinuity spacing (after Priest and Hudson [37J)

•

After the importance ratings of the classification parameters are established, the ratings for the five parameters listed in section (a) of Table 2 are summed to yield the basic (unadjusted for discontinuity orientations) Rock Mass Rating for the. structural region -under consideration. The next step is to include the sixth parameter, namely the Influence of the strike and dip orientation of discontinuities, by adjusting the basic Rock Mass Rating according to section (b) of Table 2. This step is treated separately because the influence of discontinuity orientations depends upon the engineering applications, e.g. tunnel, mine, slope or foundation. It will be noted that the 'value' of the parameter 'discontinuity orientation' is not given in quantitative terms but by qualitative descriptions such as 'favorable'. To facilitate a decision whether strike and dip orientations are favorable or not in tunneling, reference should be made to Table 4, which was originally based on studies by Wickham et al. [4J, but was modified significantly once sufficient RMR case histories were accumulated. For slopes and foundations, the reader is referred to appropriate papers [22, 23].

The parameter 'discontinuity orientation' reflects on the significance of the various discontinuity sets present in a rock mass. The main set is usually designated as set 1 and it controls the stability of an excavation, e.g. in tunneling it will be the set whose strike is parallel to the tunnel axis. The summed-up ratings of the classification parameters for this discontinuity set will constitute the overall Rock Mass Rating. On the other hand, in situations where no one discontinuity set is dominant and of critical importance, or when estimating rock mass strength and deformability, the ratings from each discontinuity set are averaged for the appropriate individual classification parameter.

In the case of civil engineering projects, an adjustment for discontinuity orientations will generally suffice. For mining applications or deep tunnels, other adjustments may be called for, such as the stress at depth or a change in stress, and these have been discussed by Laubscher [1OJ and Kendorski et al. [19]. The procedure for these adjustments is depicted in Figure 6.

•

•

i

f

I

;

•

o o

3

'"

~ 0 0

VI 0' ...,

.:;;

0-

...,

'<

o

..- Il> 0

03""

'"

~ .e.

on 0 - ..., ...,

0", on ...,

on _.

::I o

.;:;' e:..

o

z

o 0 ..., ::I

o

o A

..., ..-

o

9 I o tv

..-

00 ...,~

V>

o ...,

o tv

b

v.

tv V>

o I

..., ..-

tv V>

o ...,

v

o

..., ..-

tv V>

o o ::I

§ o·

::I o

......

0- tn· (") o

a s·

~ .

o·

on

w o

...,

o c:

A ~

3 3

(")en o 0 ::I'" _ Il>

- . ...,

::I Il> c: c, o 0 c: ::I on

o

o e OQ o

A

V>

..-

..- I 0

o v>...,

3 3

3 3

o

v

V>

o ...,

tv V

o tv

3

o 0,

..- I V> tv

3

8

I 0\

;:; 8

3 3

en o ;::>

OQ o c:

OQ o

V

A V> g 3 3

3 3

V tv V> o

..-

tv

8

I tv V> o

V> o I

.....

8

00

tv V> I V> o

tv V> I V> o

V> I tv V>

w A tv V>

.....

I V>

A

o

v

.::.. I

;:;

tv

l..

.-

I tv

•

o iii

'" '" (1)

'"

<

(1)

Q ....

~ _8

o 1

0.. 00

'"'

o

o

:-;-

o iii

'" '" (1)

'"

000

'<
wO' CJ ~
~8:: 0 00 t>
110 - 8.._0 '"
- _I 0
• '"' 0\ ..,
!JI83 0 !JI IV IV t>
(") <::r
V> :-;- ii>
"0
I'>
::l ....

o

::l"

.... 0'

t;;8~

1 1 IV IV IV!JI

!JI83

-

-

-

-

<

w
0
e. -e ~
(1)
::l '"' ..,
-e -e
• ,,"0' "0 <" ;::
-'"' < 0 <S
~8~ 0 IV gjIV .... t>
'"' 0 !JI IV '"
3 '"' c
0 ....
'" (') Q
"0 :-;- <::r
I'> ii>
::l Table 3 The RMR System: Guidelines for Classification of Discontinuity Conditions [1]

Parameter"

Ratings
<1m 1-3 m 3-10 m 10-20 m >20m
6 4 2 0
None < 0.1 mm 0.1-1.0 mm 1-5 mm >5mm
6 5 4 1 0
Very rough Rough Slightly rough Smooth Slickensided
6 5 3 1 0
Hard filling Soft filling
None <5mm >5mm -c S mm >5mm
6 4 2 2 0
Unweathered Slightly Moderately Highly Decomposed
weathered weathered weathered
6 5 3 1 0 •

Discontinuity length (persistence/continuity)

Separation (aperture)

Roughness

Infilling (gouge)

Weathering

'Some conditions are mutually exclusive. For example, if infilling is present, it is irrelevant what the roughness may be, since its effect will be overshadowed by the influence of the gouge. In such cases, use Table 2 directly.

•

Strength of
intact rock
Rating: 0-15 ~ Blasting damage
adjustmen t,Ab
0.8-1.0

I Discontinuity I
Discontinuity Orientation
density adjustment fnsitu stress and
ROD: 0-20 change of stress
Spacing:0-20 adjustment
Rating: 0-40 ~ As
.1 Basic RMR I
0.6-1.2
I 0-100


Discontinuity
condition Major faults and
Rating: 0-30 - fractures
5
0.7-1.0
l
Adjusted RMR
Groundwater
condition RMR XAb x ~s x~
Rating: 0-15 - maximum 0.5


I Support recommendations I Figure 6 Adjustments to the RMR system for mining applications

•

After the adjustment for discontinuity orientations, the rock mass is classified according to section (c) of Table 2, which groups the final (adjusted) Rock Mass Rating (RMR) into five rock mass classes, the full range of the possible RMR values varying from zero to 100. Note that the rock mass classes are in groups of 20 ratings each. This concept of rating a rock mass out of a maximum value of 100 has an advantage over an open-ended system in that it allows us to get the sense of a relative quality, or the lack of it, of a given rock mass in terms of its maximum potential.

Next, section (d) of Table 2 gives the practical meaning of each rock mass class by relating it to specific engineering problems. In the case of tunnels, chambers and mines, the output from the RMR system is the stand-up time and the maximum stable rock span for a given Rock Mass Rating, as shown in Figure 7.

1

I

l!

i i

•

Table 4 Effect of Discontinuity Strike and Dip Orientations in Tunneling (after Bieniawski [17])

Orientation of strike Dip
< 20° 20-45° 45-90°
Perpendicular to Fair Favorable Very favorable
tunnel axis: drive
with dip
Perpendicular to Fair Unfavorable Fair
tunnel axis: drive
against dip
Parallel to Fair Fair Very unfavorable
tunnel axis E

1 day 1 week 1 month

c e a. III

~

100

10'

Stand-up time Lh l

Figure 7 Relationship between the stand-up time and span for various rock mass classes according to the RMR system: output for tunneling and mining. The plotted data points represent roof falls studied: filled squares for mines, open squares for tunnels. The contour lines are limits of applicability.

When mixed quality rock conditions are encountered at the excavated rock face, such as 'good quality' and 'poor quality' being present in one exposed area, it is essential to identify the 'most critical condition' for the assessment of the rock strata. This means that the geological features which are most significant for stability purposes will have an overriding influence. For example, a fault or a shear in a high quality rock face will playa dominant role, irrespective of the high rock material strength in the surrounding strata.

It is recommended that when there are two or more clearly different zones in one rock face, one approach to adopt is to obtain RMR values for each zone and then compute the overall weighted value by the surface area corresponding to each zone in relation to the whole area, as well as by the influence that each zone has on the stability of the whole excavation.

The RMR system provides guidelines for the selection of rock reinforcements for tunnels, in accordance with Table 5. These guidelines depend on such factors as the depth below the surface (in situ stress), tunnel size and shape, and the method of excavation. Note that the support measures given in Table 5 represent the permanent and not the primary or temporary support. Approximate support guidelines, as suggested by Hoek [38], are depicted in Figure 8. Both Table 5 and Figure 8 are applicable to rock masses excavated using conventional drilling and blasting procedures.

Most recently, Lauffer [6J presented a revised stand-up time diagram specifically for tunnelboring machine (TBM) excavation and superimposed it on the RMR diagram given in Figure 7. The result is depicted in Figure 9, which is most useful because it demonstrates how the boundaries of

564

Rock Mass Classification

Support Shotcrete

Table 5 RMR System Guidelines for Excavation and Support in Rock Tunnels (after Bieniawski [17])".

•

Rock mass Excavation

class

Very good Full face, 3 m advance

rock, I

RMR: 81-100

Good rock, II RMR: 61-80

Fair rock, III RMR: 41-«)

Poor rock, IV RMR: 21-40

Very poor rock, V RMR: ~20

Full face, 1.0-1.5 m advance. Complete support 20 m from face

Top heading and bench, 1.5-3 m advance in top heading. Commence support after each blast. Complete support 10m from face

Top heading and bench, 1.0-1.5 m advance in top heading. Install support concurrently with excavation 10 m from face

Multiple drifts.

0.5-1.5 m advance in top heading. Install support concurrently with excavation. Shotcrete as soon as possible after blasting

Generally no support required except for occasional spot bolting

Rockbolts (20 mm diameter, fully grouted)

Locally bolts in crown 3 m long, spaced

2.5 m with occasional wire mesh

Systematic bolts 4 m long, spaced 1.5-2 m in crown and walls with wire mesh in crown

Systematic bolts 4-5 m long, spaced 1-1.5 m in crown and walls with wire mesh

Systematic bolts 5-6 m long, spaced 1-1.5 m in crown and walls with wire mesh. Bolt invert

Steel sets

50 mm in crown where required

None

50-100 rom in crown and 30 mm in sides

None

100-150 rom in crown and 100 mm in sides

Light to medium ribs spaced 1.5 m where required

150-200 mm in crown, 150 mm in sides and 50mm on face

Medium to heavy ribs spaced 0.75 m with steel lagging and fore-poling if required. Close invert

"Shape: horsehoe; width: 10m; vertical stress: 25 MPa; construction: drilling and blasting.

Tunneling quality index, Q
0.001 O.QI 0.1 1.0 4 10 40 100 400 1000
'" 0 0
'" 0
'" e Structurally
e U 0.1 controlled s
7n
0 failure i
~ .& !
0 0.2 f
"0 '0
c ~
:J - £ 0.3
o 0
.D !t 0' :\
Q) C
Q)"O i!:!
.~ "iii u; 0.4 0.4 •
'" <; f
'" Q)
~- .~ 0.5 0.5 ~
a. 0 '"
E :: "' (j~:;':,~m
~
0 c a.
0 E 0.6
E 0 stress
:J 0
E .2 kpz 0.7
>< - -
0 x
~ 0
'" Stress -
:::>
t induced
failure 0.9
s
• 0 1.0 1.0
0:: 0 10 20 30 40 80 90 100
I Very poor Poor Foir Good I Very good Rock mass rotings, RMR

Figure 8 Approximate support guidelines (after Hoek [38])

L-IU.).)')'LUUUfL UJ l\,ULI\. in U.).)t:.) Jut J.:,flY£fIt:t:t £fLy. .1 fit: nlYl n J Y:)Lerrt una r uiure 1 renas JOJ

10 A
• 8
E 6
c 5
0 4
a.
'"
'0 3
0
0: 2 Stand-up time (hl

Figure 9 Modified 1988 Lauffer diagram depicting boundaries of rock mass classes for TBM applications (after Lauffer [6])

RMR classes are shifted for TBM applications. Thus, an RMR adjustment can be made for

machine-excavated rock masses. .

The support load can be determined from the RMR system as proposed by Unal [20J

P = [(100 - RMR}/lOOJyB

(I)

where P is the support load (kN), B is the tunnel width (m) and y is the rock density (kg m -3).

It must be emphasized that for all applications such as those involving the selection of rock reinforcement and determination of rock loads or rock mass strength and deformability, it is the actual RMR value that must be used and not the rock mass class within which this RMR value falls. In this way, the RMR system is very sensitive to individual parameters because within one rock mass class, e.g. 'good rock', there is much difference between RMR = 80 and RMR = 61.

Finally, note that the ranges in Table 2 follow the recommendations of the International Society of Rock Mechanics (ISRM) Commission on Standardization and Classification. The interested reader is referred to an ISRM document entitled 'Suggested Methods for Quantitative Description of Discontinuities in Rock Masses' [16].

22.5.3 Applications of the RMR System

The RMR system has found wide applications in various types of engineering projects such as tunnels, slopes, foundations and mines. Most of the applications have been in the field of tunneling [15J.

This classification system has also been used widely in mining, particularly in the USA, India and Australia. Initially, Laubscher [10J applied the RMR system in asbestos mines in Africa. More recently, the RMR system was applied to coal and hard rock mining [13, 14, 19, 20, 24].

The RMR system is also applicable to slopes [22J and to rock foundations [23]. This is a useful feature which can assist with the design of slopes near the tunnel portals as well as allow estimates of the deformability of rock foundations for bridges and dams. Other special uses include applications to assess rock rippability, cuttability and cavability [1].

•

22.5.3.1 Rock mass deformability

In the case of rock foundations, knowledge of the modulus of deformability of rock masses is of prime importance. The RMR system has proved to be a useful method for estimating the in situ deformability of rock masses [25]. As shown in Figure 10, the following correlation was obtained

EM = 2RMR - 100

(2)

where EM is the in situ modulus of deformation in GPa and RMR > 50.

•

.... o

en ~

::;

"0 o E

+ tjlemaWSKI 11"'(01 to

• Serafin and Pereira (1983) /

o Stephens and Bonks (1989) Ie

+, +

++~ EM· 2 RMR -100'/1'

.'

0//

I

li+

+++0

E ·IOIRMR-IOI/40 ++/+

M \.A

0"+· ..... .

~~-; ~I

___ - .. - v-CO +

o

20 40

100

Geomechanics rock mass rating (RMRI

Figure 10 Correlation between the in situ modulus of deformation and Rock Mass Rating

Subsequently, Serafim and Pereira [21] provided many results in the range RMR < 50 and proposed a new correlation

EM = 1O(RMR-l0l/40

(3)

In the case of slopes, the output is given in section (d) of Table 2 as the cohesion and friction of the rock mass. Romana [22] has applied the RMR system extensively for the determination of rock slope stability.

22.5.3.2 Rock mass strength

Hoek and Brown [26] proposed a method for estimating rock mass strength which makes use of the RMR classification. The criterion for rock mass strength is as follows

(4)

where (J 1 is the major principal stress at failure, (J 3 is the applied minor principal stress, (J c is the uniaxial compressive strength of the rock material, and m and s are constants dependent upon the properties of the rock and the extent to which it has been fractured by being subjected to (J 1 and (J 3' For intact rock m = mj, and this is determined from a fit of the above equation to triaxial test data from laboratory specimens, taking s = 1 for rock material.

For rock masses, the constants m and s are related to the basic (unadjusted) RMR as follows [27].

For undisturbed rock masses (smooth-blasted or machine-bored excavations)

m miexp[(RMR - l(0)/28J s = exp[(RMR - l(0)/9J

(5) (6)

For disturbed rock masses (slopes or blast-damaged excavations)

m

miexp[(RMR - l(0)/14J exp[(RMR - 1(0)/6J

(7) (8)

s

•

The typical values of m and s for various rock types and corresponding to various RMR values are listed in Table 6.

Yudhbir [40] studied a rock mass criterion of the form proposed by Bieniawski [41]

a 1 [(j 3 Ja

- =A+B -

(jc (jc

(9)

•

•

8888

ci""';ci""":;

- -

N Q\ -aooMoo lIiO\O-

000~0

-

N Q\ ("'.100\(')00 oooa..._.

trioo\o

N Q\ ~oooooo -Ot"--- .......

"';0000

N Q\ MCX)V)OO "d"Ooo .......

r"io"';o

N Q\ 000000 ~O.....-l--

NO~O

Q\ 00

V)80~ ~888 oONo

Q\ 00

008r-~ ~80\8 0000

E '" E '"

-

§m~~

0000

<ri '" o:s

e

~ u o

....

where a = 0.75 and A is a function ofrock mass quality (note that A = 1 for intact rock), namely

A = exp[(RMR - 1(0)/14]

•

and B depends on rock type as follows: shale and limestone, B = 2; siltstone and mudstone, B = 3; sandstone and quartzite, B = 4; and no rite and granite, B = 5. For coal, a = 0.6 and B = 4 .

Ramamurthy [39] introduced the following relationship

(10)

where a 1 is the strength of the rock mass, a 3 is the confining stress and a em is the unconfined compressive strength of the rock mass given by

<Tern = <Teexp[(RMR - 100)/18.75]

where a c is the uniaxial compressive strength of the rock material and

Bm = Bexp[(RMR - 100)/75.5]

B depends on the rock type, namely: shale and sandstone, B = 2.2; limestone, B = 2.4; basalt, B = 2.6; marble, B = 2.8; and granite, B = 3.0. a is the slope of a line through log( a cia 3) versus log(al - (3)la3, falling within the narrow range 0.75-0.85. It is normally assumed that a = 0.8.

Moreno Tallon [29] developed a series of correlations between tunnel deformation, RMR and time, based on a case history in Spain. Unal [20] proposed an 'integrated approach' to roof stability assessment in coal mines by incorporating RMR with roof span, support pressure, time and deformation. This is diagrammatically depicted in Figure 11. Finally, recent research by Nicholson [30], incorporating the RMR system, proposed an empirical, nonlinear, stress-dependent, constitutive relationship for rock masses.

22.5.3.3 Advantages and limitations

• The RMR system is very simple to use and the classification parameters are easily obtained from

borehole data or underground mapping. This classification method is applicable and adaptable to

-,

-,

, ,

-,

p =

IOO-RMR 100

yB

Ground reaction curve

c
.S?
'\_, -0
'\" E
.2
'"
0
• Figure 11 Integration of Rock Mass Rating with support characteristics and roof deformation in coal mines (after Unal (20])

•

•

IliaIlY U111t:lt::Ul 1 V~l\.. C:;UOIU~~J.Hlb ~1\.Ua."'.lVl1':) auu. 1"-'J..1\..1...;) .I.L."'''''U "''-' \.6..., ..... "H ....... ~ .......... _.A, _'-40 __ .... -"' ..... --t"' ... ~ L- ....... ' _',

29, 30]. The method is also suitable for knowledge-based expert systems. With the application of fuzzy set methodology to the RMR system [31], the subjectiveness or fuzziness, inherent in a classification system, can be considered and incorporated into an expert system.

However, the output from the RMR classification method tends to be rather conservative, which could lead to overdesign of support systems. This aspect is best overcome by monitoring rock behavior during tunnel construction and adjusting rock classification predictions to local conditions.

Finally, the RMR system - as with any other classification system - is not to be taken as a substitute for engineering design. This classification is only a part of the empirical design approach, which is one of the three main design approaches in rock engineering (empirical, observational and analytical). It should be applied intelligently and used in conjunction with observational and analytical methods to formulate an overall design rationale compatible with the design objectives and site geology [42].

For the convenience of the user, a microcomputer program has been developed for the determination of the Rock Mass Rating and the resulting rock mass properties [1].

22.5.3.4 An example of use

The following is a typical example of finding rock mass quality by the RMR system.

Given: Rock mass in horizontal bedding, 152 m (500 ft) below surface.

Rock material: shale.

Uniaxial compressive strength: 50 MPa (5800 psi).

Strata conditions: separation < 1mm; slightly weathered, slightly rough surfaces, no infillmg.

RQD = 60%. Three discontinuity sets.

Spacing of main discontinuities: 150 mm (6 in). Groundwater conditions: damp.

In situ stresses: horizontal stress = 2.5 x vertical stress.

Solution using the RMR system .

Rating due to uniaxial compressive strength 5

Rating due to RQD 12

Rating due to discontinuity spacing 7

Rating due to condition of discontinuities 17

Rating due to groundwater conditions 10

Rating due to strike and dip orientations - 5

(Horizontal bedding = 'fair' orientation)

ROCK MASS RATING

46 0.9 41

Factor due to in situ stresses

FINAL ROCK MASS RATING, RMR (Rock mass class III: fair rock)

Rock mass modulus of deformation Cohesion of the rock mass

Friction of the rock mass

5.9 GPa 200 kPa 25°

Rock mass strength parameters for the Hoek-Brown criterion:

smooth excavation, m = 1.353 disturbed excavation, m = 0.183

s = 0.0014 s = 0.00005.

22.5.4 Database for the RMR System

The database used for the development of a rock mass classification may indicate the range of its applicability. For example, the RMR system originally involved 49 case histories (which were reanalyzed by Unal [20J), followed by 62 case histories added by Newman and Bieniawski [24] and a further 78 tunneling and mining case histories collected between 1984 and 1989. To date, the RMR

•

system has been used in 351 case histories [1]. To demonstrate the RMR database ranges, histograms are given in Figures 12, 13 and 14, depicting the ranges of the RMR values, spans of excavations, and depths below the surface applicable to the case histories on the basis of which the RMR system was developed.

It was also found that the system could be successfully used in rock formations not featured in the original case histories [33]. At the same time, in some cases, the system did not provide realistic results [32]. Nakao et al. [12] made a significant contribution by performing a statistical reconsideration of the parameters for the RMR system to assess its applicability to Japanese geological conditions. In total, 152 tunnel cases were studied. It was found that the results of the parameterrating analysis 'virtually agreed with those of the RMR concept'.

22.5.5 Correlations with Other Classification Systems

A correlation has been proposed between the RMR system and the Q index [17], and between the RMR and RSR systems [1]. Based on 111 case histories analyzed for this purpose (involving 62 Scandinavian cases, 28 South African cases and 21 case histories from the USA, Canada, Australia

80

III

5:

o u

'0

'CI> .0

E

:J Z

•

<20 21-~ 31-40 41-50 51-60 61-70 71-130 81-90 >91

RMR range (rn)

Figure 12 Distribution of RMR values in the case histories studied (after Bieniawski [1J)

120

o

o

"

'0

'CI> .0

E

:J Z

•

<3 3-4 4-5 5-7 7-10 10-15 15-20 20--25 >25

- Span range (m)

Figure 13 The range of spans encountered in the RMR case histories (after Bieniawski [1J )

0
10 0 0 0 0 0 0 0 0 ~ ~
N 10 0 10 0 10 0 10 8
v I T T N N 10 ,._
10 0 0 I I I I T
N s 0 0 ~ I I
10 2 0 8
0 10 ~ §
N (\l 2
(\l
Depth range (m)
'1 Figure 14 The range of depths in the RMR case histories (after Bieniawski [1]) 60
..
G>
..
0
<.J
'0 40
...
G>
..c
E
:::I
Z Excep<ionolly IOOr--~~--L-~---L------~--~~---L~~~~~--,

• Ngi case studies

o GeomecllOnicol case studies .. Otlle r case studies

a indion case studies

0> C

~

., .. o E

•

•

•

RMR'9InQ+44

0.1 1.0 10 100

Rock moss quality (Q)

Figure 15 Correlation between the RMR and the Q index (after Jeghwa [36])

and Europe), the results are plotted in Figure 15 from which the following relationship is found to be applicable for tunnels

RMR = 9lnQ + 44

(11)

For mining drifts, Abad et al. [3).] analyzed 187 coal mine roadways in Spain, arriving at the correlation

RMR = 10.5 In Q + 42

Rutledge [11] determined from seven tunneling projects the following correlation

(12)

RSR = O.77RMR + 12.4

(13)

18. Weaver J. Geological factors significant in the assessment of rippability. Civil Engineer in South Africa 17, 313-316 (1975).

19. Kendorski F., Cummings R., Bieniawski Z. T. and Skinner E. Rock mass classification for block caving mine drift support. In Proc. 5th Int. Congr. Rock Mech., Melbourne, pp. B51-B63. Balkema, Rotterdam (1983).

20. Unal E. Design Guidelines and Roof Control Standards for Coal Mine Roofs. Ph.D. Thesis, p. 355. Pennsylvania State University (1983).

21. Serafim J. L. and Pereira J. P. Considerations of the Geomechanics Classification of Bieniawski. In Proc. Int. Symp, Engineering Geology and Underground Construction, Lisbon, vol 1, pp. II.33-1I.42. LNEC, Lisbon (1983).

22. Romana M. New adjustment ratings for application of Bieniawski classification to slopes. In Proc. Int. Symp. Rock Mechanics in Excavationsfor Mining and Civil Works, Mexico City, pp. 59-68. ISRM (1985).

23. Bieniawski Z. T. and Orr C. M. Rapid site appraisal for dam foundations by the Geomechanics Classification. In Proc. 12th Congr. on Large Dams, Mexico City, pp. 483-501. ICOLD (1976).

24. Newman D. A. and Bieniawski Z. T. Modified version of the Geomechanics Classification for entry design in underground coal mines. Trans. Soc. Min. Eng. AIME 280, 2134-2138 (1986).

25. Bieniawski Z. T. Determining rock mass deformability: experience from case histories. Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. 15, 237-247 (1978).

26. Hoek E. and Brown E. T. Empirical strength criterion for rock masses. J. Geotech. Eng. Div., Am. Soc. Civ. Eng. 106, 1030--1035 (1980).

27. Hoek E. and Brown E. T. The Hoek-Brown failure criterion - a 1988 update. In Proc. 15th Can. Rock Mech. Symp.

University of Toronto (1988).

28. Stacey T. R. and Page C. H. Practical Handbook for Underground Rock Mechanics, p. 15. Trans Tech, Clausthal, Germany (1989).

29. Moreno Tallon E. Comparison and application of the Geomechanics Classification schemes in tunnel construction, In Proc. Tunneling '82, pp. 241-246. Institution of Mining and Metallurgy, London (1982).

30. Nicholson G. A. and Bieniawski Z. T. An empirical constitutive relationship for rock mass. In Proc. 27th U.S. Symp.

Rock Mech. Tuscazoosa, AL (Edited by H. L. Hartman), pp. 760--766. AIME, New York (1986).

31. Nguyen V. U. and Ashworth E. Rock mass classification by fuzzy sets. In Proc. 26th U.S. Symp. Rock Mech., Rapid City, SD (Edited by E. Ashworth), pp. 937-946. Balkema, Rotterdam (1985).

32. Kaiser P. K., MacKay C. and Gale A. D. Evaluation of rock classifications at B. C. Rail Tumbler Ridge tunnels, Rock Mech. Rock Eng. 19, 205-234 (1986).

33. Singh R. N., Elmherig A. M. and Sunu M. Z. Application of rock mass characterization to the stability assessment and blast design in hard rock surface, mining excavations. In Proc. 27th U.S. Symp. Rock Mech., Tuscazoosa, AL (Edited by H. L. Hartman), pp. 471-478. AIME, New York (1986).

34. Goodman R. E. Introduction to Rock Mechanics, 2nd edn., pp. 42-49. Wiley, New York (1989).

35. Abad J., Celada B., Chacon E., Gutierrez V. and Hidalgo E. Application of Geomechanical Classification to predict the convergence of coal mine galleries and to design their supports. In Proc. 5th Int. Congr. Rock Mech., vol. 2, Melbourne, pp. EI5-EI9. Balkema, Rotterdam (1983).

36. Jethwa J. L., Dube A. K., Singh B. and Mithal R. S. Evaluation of methods for tunnel support design in squeezing rock conditions. In Proc. 4th Int. Congr. International Association of Engineering Geology, vol. 5, pp. 125-134. Balkema, Rotterdam (1982).

37. Priest S. D. and Hudson J. A. Discontinuity spacings in rock. Int. J. Rock Mech. Min. Sci. & Geomech. Abstr. 13, 135-148 (1976).

38. Hoek E. Geotechnical design oflarge openings at depth. In Proc. Con! Rapid Excavation and Tunneling, pp. 1167-1180.

AIME, New York (1981).

39. Ramamurthy T. Stability of rock mass. Indian Geotech. J. 16, 1-74 (1986).

40. Yudhbir E. T. An empirical failure criterion for rock masses. In Proc. 5th Int. Congr. Rock Mech., vol. 1, Melbourne, pp. BI-B8. Balkema, Rotterdam (1983).

41. Bieniawski Z. T. Estimating the strength of rock materials. J. S. Afr. lnst: Min. Metall. 74, 312-320 (1974).

42. Bieniawski Z. T. Design Methodology in Rock Engineering; Balkema, Rotterdam (1992).

II ii'

! '

:!

:i

I

\

Weaver 1. Geological factors significant in the assessment of rippability. Civil Engineer in South Africa 17, 313-316 ~~~~~rski F., Cummings R., Bieniawski Z. T. and Skinner E. Rock mass classification for block caving mine drift support. In Proc. 5th Int. Congr. Rock Mech., Melbourne, pp. B51-B63. Balkema, Rotterdam (1983).

Unal E. Design Guidelines and Roof Control Standards for Coal Mine Roofs. Ph.D. Thesis, p. 355. Pennsylvania State University (1983).

Serafim J. L. and Pereira J. P. Considerations of the Geomechanics Classification of Bieniawski. In Proc. Int. Symp. Engineering Geology and Underground Construction, Lisbon, vol 1, pp. 11.33-11.42. LNEC, Lisbon (1983).

Romana M. New adjustment ratings for application of Bieniawski classification to slopes. In Proc. Int. Symp. Rock MechaniCS in Excavations for Mining and Civil Works, Mexico City, pp. 59-68. ISRM (1985).

Bieniawski Z. T. and Orr C. M. Rapid site appraisal for dam foundations by the Geomechanics Classification. In Proc. 12th Congr. on Large Dams, Mexico City, pp. 483-501. ICOLD (1976).

Newman D. A. and Bieniawski Z. T. Modified version of the Geomechanics Classification for entry design in underground coal mines. Trans. Soc. Min. Eng. AIME 280, 2134-2138 (1986).

Bieniawski Z. T. Determining rock mass deformability: experience from case histories. Int. 1. Rock Mech. Min. Sci. &: Geomech. Abstr. 15,237-247 (1978).

Hoek E. and Brown E. T. Empirical strength criterion for rock masses. 1. Geotech. Eng. Div., Am. Soc. Civ. Eng. 106, 1030-1035 (1980).

Hoek E. and Brown E. T. The Hoek-Brown failure criterion - a 1988 update. In Proc. 15th Can. Rock Mech. Symp. University of Toronto (1988).

Stacey T. R. and Page C. H. Practical Handbook for Underground Rock Mechanics, p. 15. Trans Tech, Clausthal, Germany (1989) .

. Moreno Tallon E. Comparison and application of the Geomechanics Classification schemes in tunnel construction, In Proc. Tunneling '82, pp. 241-246. Institution of Mining and Metallurgy, London (1982).

Nicholson G. A. and Bieniawski Z. T. An empirical constitutive relationship for rock mass. In Proc. 27th U.S. Symp. Rock Mech. Tuscazoosa, AL (Edited by H. L. Hartman), pp. 760-766. AIME, New York (1986).

Nguyen V. U. and Ashworth E. Rock mass classification by fuzzy sets. In Proc. 26th U.S. Symp. Rock Mech., Rapid City, SD (Edited by E. Ashworth), pp. 937-946. Balkema, Rotterdam (1985).

Kaiser P. K., MacKay C. and Gale A. D. Evaluation of rock classifications at B. C. Rail Tumbler Ridge tunnels, Rock Mech. Rock Eng. 19, 205-234 (1986).

Singh R. N., Elmherig A. M. and Sunu M. Z. Application of rock mass characterization to the stability assessment and blast design in hard rock surfacemining excavations. In Proc. 27th U.S. Symp. Rock Mech., Tuscazoosa, AL (Edited by H. L. Hartman), pp. 471-478. AIME, New York (1986).

Goodman R. E. Introduction to Rock Mechanics, 2nd edn., pp. 42-49. Wiley, New York (1989).

Abad J., Celada B., Chacon E., Gutierrez V. and Hidalgo E. Application of Geomechanical Classification to predict the convergence of coal mine galleries and to design their supports. In Proc. 5th Int. Congr. Rock Mech., vol. 2, Melbourne, pp. EI5-EI9. Balkema, Rotterdam (1983).

Jethwa 1. L., Dube A. K., Singh B. and Mithal R. S. Evaluation of methods for tunnel support design in squeezing rock conditions. In Proc. 4th Int. Congr. International Association of Engineering Geology, vol. 5, pp. 125-134. Balkema, Rotterdam (1982).

Priest S. D. and Hudson J. A. Discontinuity spacings in rock. Int. 1. Rock Mech. Min. Sci. & Geomech. Abstr. 13, 148 (1976).

Hoek E. Geotechnical design oflarge openings at depth. In Proc. Conf Rapid Excavation and Tunneling, pp. 1167-1180. AIME, New York (1981).

Ramamurthy T. Stability of rock mass. Indian Geotech. 1. 16, 1-74 (1986).

Yudhbir E. T. An empirical failure criterion for rock masses. In Proc. 5th Int. Congr. Rock Mech., vol. 1, Melbourne, pp. BI-B8. Balkema, Rotterdam (1983).

Bieniawski Z. T. Estimating the strength of rock materials. 1. S. Afr. Inst. Min. Metall. 74, 312-320 (1974). Bieniawski Z. T. Design Methodology in Rock Engineering. Balkema, Rotterdam (1992).

COMPREHENSIVE ROCK ENGINEERING

Principles, Practice & Projects

Editor-in-Chief JOHN A. HUDSON

Imperial College of Science, Technology & Medicine, London, UK

Volume 3

ROCK TESTING AND SITE CHARACTERIZATION

•. 1

"

,it:

"'"

:1

Volume Editor JOHN A. HUDSON

Imperial College of Science, Technology & Medicine, London, UK

PERGAMON PRESS

OXFORD' NEW YORK· SEOUL' TOKYO