Denkhaus (1970) pointed out the existence of a gap between the acquisition of rock mechanics data and the final decisions of engineers. How can engineers best use the information given to them by geologists, geophysicists and specialists in rock mechanics ? What type of information is required and by whom? The consulting engineer wants to know how rock will behave over a long period of time, what strains it will develop, what stresses it will withstand and what load will be transmitted to the structures. The contractor is mainly concerned with short-term problems, with rock stability and rock behaviour during the different phases of the construction work. The planning and the progress of the construction will depend on rock properties in which the consulting engineer may be less interested. Major problems concerning legal responsibilities and price levels may depend on different interpretations of rock characteristics.
(a) Before discussing rock classifications established by geophysicists and rock mechanicists, it is necessary to mention the approach to such a problem
Engineering classification of jointed rock masses 155 by geologists. The useful work of Duncan, Deere and others has been dealt with in sections 2.1.1, 4.1.2 and 6.6.1. Dearman (1974) has recently published an interesting paper summarizing a geologist's approach to engineering rock classifications, based on his experience with geology in the British Isles. This author produced a table detailing a classification of igneous rocks on the basis of mineralogy and grain size (a classical approach to be found in many textbooks). Other tables give similar classifications of sedimentary rocks.
He then summarizes the work of Duncan, Jones, Deere, etc., which has
Rock mass classification
1 200 MPa 100 MPa
•§ 50 MPa o 25 MPa
10 MPa Solid (Almost no
joints)
Massive (Little jointed) Strong rock mass Cohesion: >0.2MPaoti
Blocky/seamy (Moderately
jointed)
Medium-strength rock mass Cohesion: 0.1-0.2 MPa or friction: 30°'
Weak rock mass
Cohesion: 10-400 kPa or friction: 20^30°
Very weak rock mass
Fractured (Intensely jointed)
10 m 3 m 1 m 300 mm Spacing of joints
Crushed and shattered
Cohesion < lOkPa Friction < 20°
100 mm 50 mm 10 mm
Fig. 6.50 Strength diagram of jointed rock masses (modified from Miiller).
already been amply discussed in the preceding chapters. Such a geologist's classification could be consulted by engineers, since it is based on modern geophysical information.
(b) One of the first attempts by a geophysicist to systematically describe characteristics in order to establish an engineering classification of jointed rock masses is a well-known diagram prepared by Muller-Salzburg, repro- duced as fig. 6.50.
Miiller groups the rock masses in four classes: strong, medium strength, weak and very weak. He uses two main rock characteristics, combining the uniaxial compressive strength of intact rock and the joint spacing. Addition- ally there is mention on his diagram of the rock cohesion or of the angle of friction which can be expected for each class of rock.
(c) Geophysicists have been impressed with the advantages of the rock quality designation (RQD) proposed by Deere (1963-68) (see section 2.1.4).
In recent times some experts have felt that it would be useful to supplement the information given by the usual description of general area and of local site geology and the RQD index with additional information on joints and faults and other rock properties.
A considerable amount of work has been done in analysing hundreds of cases, by a comparison of rock characteristics and the behaviour of the engineering structures (Coates, 1964; Cecil, 1970a, b; Wickham, 1972;
Tiedmann; Skinner; etc.)-the International Society for Rock Mechanics and the International Association of Engineering Geology have both appointed Commissions to study rock classifications.
Bieniawski (1973) states that while a rock classification of jointed rock masses, based on the inherent properties of the rock mass itself, should be capable of application to practical engineering problems, it should be general enough so that the same rock would always be classified in the same way, regardless of how it is being used. The classification he suggests was developed from an analysis of many earlier classifications. It incorporates the following parameters:
(1) Rock quality designation (RQD).
(2) State of weathering.
(3) Uniaxial compression strength of intact rock.
(4) Spacing of joints or bedding.
(5) Strike and dip of joints, bedding.
(6) Openness of joints.
(7) Continuity of joints.
(8) Ground water inflow.
Based on these parameters, an engineering classification of jointed rock masses, termed the 'Geomechanics Classification' is proposed in table 6.4.
For the purpose of this classification, it is necessary to divide the rock mass into a number of domains, each having similar structural characteristics; for example the same rock type or the same joint spacing. The boundaries of a domain may coincide with such geological features as faults or dykes.
Experience shows how rapidly rock characteristics change when crossing contact zones.
For simple classification, rock masses are grouped into only five classes.
The various parameters need not necessarily conform to the same rock class:
for example, a rock with an RQD of 70 % (class 3) may have a joint spacing of 0-2 m (class 4) and display no ground-water inflow (class 2).
(1) It is easy to establish five classes for the first parameter, the RQD index varying from 0 to 25%, 25 to 50%, etc.
(2) For the second parameter, rock weathering, Bieniawski suggests the following classes:
Unweathered: no visible sign of weathering.
Slightly weathered: penetrating weathering developed on open discon- tinuity surfaces, slight weathering of rock material.
Moderately weathered: slight discoloration extends through the greater part of the rock mass.
Item
1 2 3 4 5 6 7 8
Table 6.4 Geomechanics classification of jointed rock Class No and its description
Rock quality RQD (%) Weathering
Intact rock strength, MPa Spacing of joints
Separation of joints Continuity of joints Ground-water inflow
(per 10 m of adit) Strike and dip orientations
1 Very good
90-100 Unweathered
>200
> 3 m
< 0 1 mm Not continuous
None Very favourable
2 Good 75-90 Slightly weathered
100-200 1-3 m
< 0 1 mm Not continuous
None Favourable
masses (after 3 Fair 50-75 Moderately
weathered 50-100 0-3-1 m 0-1-1 mm Continuous,
no gouge Slight <25
litres/min Fair
Bieniawski) 4 Poor 25-50 Highly weathered
25-50 50-300 mm
1-5 mm Continuous,
with gouge Moderate 25-125
litres/min Unfavourable
5 Very poor
<25 Completely
weathered
<25
<50mm
> 5 mm Continuous,
with gouge Heavy >125
litres/min Very unfavourable
Highly weathered, weathering extends throughout rock mass, rock material partly friable.
Completely weathered: rock totally discoloured and decomposed.
(3) The third parameter concerns the uniaxial compressive strength of intact rock. The rocks can be grouped in five classes (1 MPa = 10-2 kg/cm2).
Very low strength, 1-25 MPa: chalk, rock salt.
Low strength, 25-50 MPa: coal, saltstone, schist.
Medium strength, 50-100 MPa: sandstone, slate, shale.
High strength, 100-200 MPa: marble, granite, gneiss.
Very high strength > 200 MPa quartzite, dolorite, gabbro, basalt.
(4) and (5) Five classes are chosen for the spacing of joints, varying from class 5, for distances smaller than 50 mm to class 1, when the distance is larger than 3 m (as can be seen in table 6.4). Similarly the separation of joints is assumed to vary between larger than 5 mm (class 5) to smaller than 0-1 mm (class 1).
(6) to (8) These parameters are classified as shown in table 6.4.
It is advantageous to assign a rating to each parameter by a weighted numerical value. The final rock class rating will be the sum of weighted values determined for the individual parameters, higher numbers reflecting better conditions and hence lesser support in the case of tunnels. Based on a study of Wickham et ah the importance of rating as shown in table 6.5 is proposed by Bieniawski.
It should be noted, for example, that items 1 and 2, representing data obtainable from cores, receive 25% rating, while the intact rock strength (item 3) is worth 10%. Items 4 and 5, representing field data on joints, receive as much as 45 %.
Bieniawski has developed some further aspects of this 'Geomechanics Classification' concerning tunnel design and construction. These will be discussed in section 10.9, together with further research work on similar problems.
The eight parameters introduced by Bieniawski allow a fair description of jointed rock masses, but there are some limitations. According to Bieniawski 'special caution should be exercised in the case of shales and other swelling materials. These rocks are characterized by a wide variation in their engineer- ing properties, particularly their durability (resistance to weathering) under conditions of wetting and drying' (see Franklin, 1972 and Olivier, 1973). A similar remark could be made for rock masses with a tendency to slow creep.
A Norwegian team (Barton et ah, 1974) objects that Bieniawski has almost ignored three important properties of rock masses, namely the roughness of
Table 6.5 Importance ratings (a) Individual ratings for classification parameters
Item 1 2 3 4 5 6 7
8
(b) Total ratings
Parameter
Rock quality RQD Weathering Intact rock strength Spacing of joints Separation of joints Continuity of joints Ground water
Strike and dip orientations
for rock mass classes Class no.
Description of Class Total rating
[Tunnels (Foundations
1 Very good rock
90-100 1 16 9 10 30 5 5 10
15 15
2 Good rock
70-90 2 14 7 5 25 5 5 10
13 13
Class 3 12 5 2 20 4 3 8
10 10
3 Fair rock
50-70
4 7 3 1 10 3 0 5
5 0
4 Poor rock
25-50
5 3 1 0 5 1 0 2
3 - 1 0
5 Very poor rock
< 2 5
joints, the frictional strength of joint fillings and the rock load. The impor- tance of the natural residual stresses in the rock mass will be demonstrated at some length in section 10.9.
Important cases exist where a precise, systematic description of the rock mass characteristics, along the lines suggested by Bieniawski or others, fails to detect vital causes of potential failure of the structure. In part four, several case histories will be discussed where a combination of factors, some of them unforeseen at the design stage, caused major troubles or even the collapse of structures, the description of the rock masses having failed to disclose inherent weaknesses. The cases of the collapse of the Malpasset dam, of the rock falls at Kariba North Bank Machine Hall excavation and the great Vajont rock slide will illustrate this point. Similar cases were men- tioned by Barbier (1974); final decisions on where to locate dams were taken on the basis of geological exploration at large, rather than local geology, almost disregarding rock characteristics.
7 Mathematical approach to strain-stress distribution in rock masses