Sie sind auf Seite 1von 15

PERGAMON International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

Development of an assessment system for the blastability of rock


masses
J.-P. Latham, Ping Lu
Geomaterials Unit, Queen Mary and West®eld College, London E1 4NS, U.K.
Accepted 13 September 1998

Abstract

Intact rock properties and the discontinuity structure of a rock mass are among the most important variables in¯uencing
blasting results. This in¯uence is considered to be a composite intrinsic property of a rock mass and is referred to as the
blastability of a rock mass. It represents the ease with which a rock mass can be fragmented by blasting. This paper outlines an
energy-block-transition model, recently proposed by the authors for characterising the blast process. A preliminary validation of
this model using two sets of ®eld data from the literature is brie¯y outlined. The model is comparable or better than Bond's
comminution theory at predicting blasting results for cases where intrinsic rock properties are relatively constant. To generate a
predictive capability for the model, a blastability designation BD, is designed which re¯ects the intrinsic resistance of the rock
mass to blasting. The quanti®cation of BD, based on rock engineering systems approaches and consideration of a
comprehensive range of intact rock properties and discontinuity structures is illustrated. A case study is given which applies the
model and the associated assessment system to a highway cutting site. Con®dence as to the potential value of the assessment
system and the model is obtained since re®nement and improvement on pre-existing models can be seen from the new
preliminary results. # 1999. Published by Elsevier Science Ltd. All rights reserved.

1. Introduction e€ect, until recent times, has been that blasting design
has relied on rules of thumb obtained by precedent
Blasting is the most frequently used means for quar- practice [10]. The failure to promote blast design tools
rying, mining and highway rock excavation. A blasting beyond rules of thumb might have resulted from the
operation can be comprehensively described by: intact fact that the in¯uence of in-situ rock properties, dis-
rock and rock mass properties (the concern of this continuity structures and their interactions are often
paper), explosive properties, blasting geometry or pat- too dicult to be quantitatively isolated and identi®ed.
tern and initiation sequences, etc. The rock engineering systems methodology developed
The in¯uence of intact rock and rock mass proper- by Hudson [18], aims to provide both a useful check-
ties on blasting operations has long been studied [1± list for the in¯uential factors of rock engineering pro-
12]. This in¯uence has been mentioned and incorpor- jects and a logical framework for the complete design
ated in various ways, such as Bond's work index [13], procedure. As such, it has potential for coping with
Hino's blastability coecient [14], rock factor [15] and complex rock engineering problems, such as the re-
blastability index [16]. However, little attempt has been sponse to blasting.
made to develop a quantitative parameter or system to It was noted from ®ndings reported by di€erent
de®ne the ease of fragmentation of rock by blasting, in researches [4±6, 16, 19] and the present authors [9, 20±
spite of the fact that this kind of development was 22] that in-situ rock mass properties are among the
suggested long ago [2] and was recently most important contributory factors in fragmentation
reemphasised [10, 17]. and that the characterisation of the blastability has
Selecting one or more parameters for the rock prop- become a pressing task for blasting operations. It was
erties that will re¯ect the resistance of the rock mass to also found that a coherent but essentially empirical
fragmentation by blasting has been a major obstacle to rock blastability system which incorporates the mech-
the description of the ease of fragmentation. The anical properties of a rock mass, the in-situ block size

0148-9062/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 8 - 9 0 6 2 ( 9 8 ) 0 0 1 7 5 - 2
42 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

and the input energy, based on the current scienti®c


understanding of fragmentation, has not yet been
achieved. This has hindered the achievement of more
optimal blasting operations.
Consequently, a blastability assessment system has
been developed to accommodate the di€erent degree of
in¯uence on blastability of a variety of intact rock and
rock mass properties, helping to put blasting oper-
ations on a more logical footing. The development of
the blastability system follows from current under-
standing of the fragmentation process drawn from the
literature. The paper begins with the characterisation
of blastability and its physical background with respect
to energy consideration. Rock engineering systems
Fig. 1. The concept of blastability: two di€erent rock masses with
approaches are then used to establish a blastability sys- the same IBSD but with di€erent blastability are transformed to two
tem. Appropriate sources of practical data are ®nally di€erent BBSD curves (see text for explanation).
introduced for an examination of the system.
subject to identical blast design are represented by DA1
and DA2. The second rock mass proves to be intrinsi-
2. Blastability and its characterisation cally more dicult to fragment by blasting than the
®rst one, since the second blastpile contains more large
Two di€erent rock masses, when subjected to identi- blocks than the ®rst although their IBSDs are identi-
cal blast geometry and energy input from explosives, cal. The area DA bounded by the IBSD and BBSD
will produce quite di€erent degrees of fragmentation. curves and the 0 and 100% passing line, for a particu-
This is because the rock masses have inherently di€er- lar blasting operation, is considered to have the follow-
ent resistance to fragmentation by blasting. That is, ing special signi®cance: the di€erence between DA1 and
the two rock masses have a di€erent ease with which DA2 indicates the di€erence in blastability between the
they can be fragmented by blasting. This property is two rock masses. Let Es represent the explosive energy
hereafter referred to as the ``blastability'' of a rock input per unit rock mass that is consumed in trans-
mass. It appears to be a kind of intrinsic property, like forming the rock mass with a given IBSD into a blast-
the hardness of a rock and apart from the possibility
pile with a given BBSD, in order that a working
of blast induced fragmentation from previous rounds,
hypothesis, EsADA, can be introduced at this
it is uncontrollable. As can be seen in Section 3, many
point [22, 23].
factors a€ect the blastability of rock masses and it is
For each transformation with a certain value of DA,
therefore helpful to consider the blastability of the
an e€ective size parameter Xo is introduced to re¯ect
rock mass to be a composite intrinsic property of the
an inverse size e€ect associated with an increase in
rock mass.
energy consumption, where Xo is the value of the
2.1. The energy-block-transition model abscissa of the centre of gravity of the geometric shape
of DA. Such an inverse size e€ect for the dependence
The fragment size is a fundamental characteristic, of energy on an objective size can be illustrated as fol-
and is mainly governed by the geomechanical nature lows. A given value of DA can be associated with any
of the host rock mass. Blasting is looked upon as a transformation. For example, the transformation may
transformation from the state with an in-situ block size be for relatively large in-situ blocks to slightly smaller
distribution (IBSD) to the state with a blasted block blocks in a blastpile, and this transformation would be
size distribution (BBSD) (e.g. from IBSD-C into either associated with a relatively larger Xo. Alternatively, a
BBSD-1 or BBSD-2 in Fig. 1). This transforming pro- transformation with the same DA could be for rela-
cess is implemented by inputting a certain energy, i.e. tively small in-situ blocks being reduced to a blastpile
by detonating a quantity of explosive. The transform- resembling powdery ®nes, i.e. with a relatively smaller
ation result can be indicated by the block size distri- Xo. But, because the latter case requires the generation
bution after blasting. In Fig. 1, two di€erent rock of much more fracturing and surface area, it is logical
masses are considered which have the same IBSD, that it would consume more energy despite the fact
labelled IBSD-C. BBSD-1 and BBSD-2 are the BBSDs that the two transformations have the same DA. The
of the ®rst and the second rock masses obtained by empirical energy±size relation of other researchers
inputting an identical amount of explosive energy. The refers to an objective size parameter [24±26] that has a
transformation areas for the two di€erent rock masses similar signi®cance to this e€ective size parameter and
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 43

indicates that in general, the energy consumed in size block sizes, characterised by the mean block size Sai, is
reduction is inversely proportional to the objective size transformed into blasted blocks with small block sizes
raised to some power. Based on the above working hy- characterised by the mean block size Sab. The larger
pothesis and the de®nition of Xo, the following re- the value of Bi of a rock mass, the easier it can be
lationship is proposed by Lu [22, 23], fragmented by a given energy input for blasting.
Eqs. (4a)±(b) indicate that a rock mass with a larger Bi
DA
Es A : …1† will be easier to break down to small blocks than a
X 1=2
o rock with a lower Bi. That is, the larger the Bi, the
greater the blastability.
Detailed derivations of DA and Xo are given
elsewhere [22, 23]. They yield the following results: 2.2. Preliminary examination of the energy-block-
DA ˆ Sai ÿ Sab …2a† transition model

Sai ‡ Sab To see whether the E-B-T model proposed is an


Xo 0 , …2b† improvement on existing models requires an analysis
2
of results from practical blasting operations. A set of
where, Sai and Sab are the mean sizes of IBSD and model scale blasts, or preferably, a series of ®eld blasts
BBSD, respectively. They are given by the following with constant and given in-situ conditions, but with
mathematical de®nition for the mean of a size distri- di€erences in both blasting patterns and energy input,
bution: would be ideal for examination of the applicability of
… Su the model.
Sa ˆ Sf …S †dS, …3† So far, there have been three theories concerning the
Sl
energy associated with size reduction. These are
where, f(S) represents the distribution function of Rittinger's ®rst comminution theory, Kick's second
block size, i.e. the probability density function; Sl rep- comminution theory and Bond's third comminution
resents the lower boundary of block size, which may theory. Bond's theory is a compromise between
often be zero and Su is the upper boundary of block Rittinger's and Kick's theories and is generally recog-
size. nised to be the best model to describe blasting
Consequently, the new relationship has been operations [6]. Bond's third theory has the following
proposed [22, 23] as follows: form:
 
Sai ÿ Sab 1 1
Es ˆ …4a† Es ˆ 10Ec p ÿ p , …5†
Bi ……Sai ‡ Sab †=2†0:5 Sb80 Si80

or where, Es is the required energy for fragmentation in


kWh per ton of processing rock material; Ec is Bond's
Sai ÿ Sab work index; Sb80 and Si80 are the blastpile and in-situ
Bi ˆ , …4b†
Es ……Sai ‡ Sab †=2†0:5 block size. Subscripts 80 means that the block size is
equivalent to the sieve opening (in microns) through
where Bi is a coecient, relating the speci®c energy which 80% of the rock materials pass.
required to implement the transformation process. If the relationship between energy input and the
Eqs. (4a)±(b) relate the energy input, the character- block size before and after blasting is better described
istic sizes before and after blasting and the size distri- by Eq. (4a) than by Bond's third theory, the Bi in the
butions of the in-situ and blasted blocks. It appears E-B-T model should exhibit less variability than the Ec
that the value of Bi is an indicator of the ease with in Bond's theory. Thus, a preliminary validation has
which the process of transforming the rock mass with been carried out by comparing the extent to which the
IBSD into the blastpile with BBSD is implemented. E-B-T coecient Bi and Bond's work index, Ec diverge
The coecient Bi is herewith referred to as the energy- from a constant.
block-transition (E-B-T) coecient and Eqs. (4a)±(b) Two main sources of practical data which include
are referred to as the energy-block-transition (E-B-T) information about IBSD, BBSD and the energy input,
model. were found in the literature and were used for the pre-
The E-B-T coecient Bi is a measure of the intensity liminary validation. This was implemented by looking
of the transformation of mean block size compared to at the values of Dmax/M and sample coecient of
the objective size Xo associated with the transform- variation s/M, where Dmax represents the maximum
ation process for a given input of energy. In the trans- di€erence between the values of Bi or Ec, s and M are
formation process, the in-situ rock mass with large the standard deviation and the mean of Bi or Ec. The
44 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

Table 1
Comparison between the relative dispersion of the E-B-T coecient, Bi and that of Bond's work index, Ec for the blast results for two case stu-
dies from the literature

Case E-B-T coecient Bi (m0.5/kWh/t) Bond's work index Ec (kWh/t) Blast rounds

Dmax/M (%) s/M (%) Dmax/M (%) s/M (%)

Sandside quarry* 40.38 (4.56) 16.88 (2.14) 42.73 (30.08) 16.72 (16.23) 4 (3)
ENUSA Mine 86.57 32.72 106.18 47.06 5
Reocin Mine 31.72 12.40 58.90 23.85 4
q
Pn 2
Dmax = Xmax ÿ Xmin; and s= j ……Xj ÿ Xa † †=N. *For the Sandside quarry case, ®gures in brackets are the outcome excluding the ®rst of
the four blast rounds. This data trimming may be justi®ed on grounds that there existed a signi®cant di€erence in the stemming length, decking
and distribution of explosives between the ®rst and the other three blast rounds.

examination results are brie¯y described below and the the rock mass of interest. It will be most advantageous
details can be referred to elsewhere [22, 23]. for the coecient Bi to be determined before blasting
The ®rst case study cited is the work reported by in order to help with the blast design of an excavation
Wang [27], in which ®ve rounds of full scale blast trials operation. Without any realistic chance in the short
were carried out at the Sandside limestone quarry term of a practical analytical solution to de®ne the
located in Cumbria, England. The information about value of Bi for a given rock mass as a function of ma-
IBSD, BBSD, explosive energy input and blasting pat- terial properties, the development of a comprehensive
terns were either given or can be derived. Four out of assessment system for quantifying the blastability of
®ve blast rounds provide a data set of full-scale blast- rock masses would appear to have great potential. The
ing parameters with approximately the same geological objective would then be to have a method for deter-
in-situ conditions, but with di€erences in both blasting mining the value of Bi from results derived from the
patterns and energy input. The second case cited is the assessment system.
work reported by Aler et al. [28, 29], in which the as- In reviewing the blasting practices and literature
sociated data input from production blasts carried out published [2, 5, 6, 9, 10, 14±17, 30±36] it is obvious that
in two mines were provided. Two groups of data, one many factors can a€ect the blastability of rock masses.
from Bench 4 in ENUSA mine and another from These cited factors consist of a wide, possibly compre-
Bench 3 in Reocin Open Pit are used for checking the hensive range of intact rock properties and discontinu-
validity of the E-B-T model. ity structures, each of which in¯uences the blasting
The E-B-T coecients Bi and the Bond work indices result to a varying degree. The blastability of the rock
Ec were calculated and the relative dispersions were mass is therefore considered to be a composite intrinsic
analysed and summarised as shown in Table 1. It is property of the rock mass.
seen from Table 1 that both the coecient of variation The factors in¯uencing blasting results fall into two
and the value of Dmax/M for Bi are signi®cantly lower groups. The ®rst group is the intact rock properties,
than for Ec. This indicates that Bond's work index, Ec which includes strength, hardness, elasticity, deform-
deviates more from a constant than the E-B-T coe- ability, density of rock, etc. They are dependent upon
cient Bi and suggests that the E-B-T model gives a clo- rock texture, internal bonds, composition and distri-
ser ®t to the blast data than the alternative Bond's bution of minerals forming the rock. The second
model. group is the discontinuity structure that consists of
Thus, although further examinations of the applica- orientation, spacing and extent of discontinuities, and
bility of the E-B-T model to practical blasting oper- the in-situ block sizes, created by a range of long-term
ations are desirable, the attempted validation of the geological processes.
model using the above two cases provides sucient The problem of obtaining a satisfactory measure of
encouragement to examine further the E-B-T model blastability from an assessment of numerous poten-
and ways of quantifying Bi for di€erent rock masses. tially in¯uential factors has at least three features
which have often been neglected in early attempts to
investigate blastability. One is the interactions between
3. Factors in¯uencing blastability of rock masses factors. Another is the degree of in¯uence (or the
weighing) to be attributed to each factor or coupled
As shown above, the E-B-T coecient Bi, is a quan- factors. A third is the need to treat subjective data, a
titative measure of the blastability of a rock mass. To situation often encountered in geotechnical engineering
apply the E-B-T model to a practical blasting oper- with systems of soils, rocks, ¯uids and discontinuities.
ation requires prior knowledge of the Bi coecient for Because of the complexity of the geotechnical system,
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 45

it is often necessary to take advantage of all available could provide a better alternative assessment of blast-
data, whether it is objective observation, practical ex- ability.
perience or subjective data based on combinations of
observation, practical experience and engineering jud-
gement. The degree of in¯uence of individual factors 4. Development of an assessment system of blastability
upon each other and upon the blastability is often
expressed in terms of ``strong'', ``fair'' and ``weak'', 4.1. Rock engineering systems and interaction matrix
and it is important to be able to integrate such valu-
able knowledge. 4.1.1. Rock engineering systems
While uniaxial compressive strength might be an im- The rock engineering system (RES) aims to provide
portant indicator of the blastability of a rock, several a useful checklist for a rock engineering project. More
others such as sonic velocity or joint spacing might be importantly, it also aims to provide a framework from
of equal or greater importance, especially when which the complete design procedure can be evaluated,
coupled in a certain way. It is perhaps now more leading a rock engineering project to an optimal result.
apparent, considering the three factors mentioned An RES description of the overall interactive mechan-
above, why blastability assessment presents such a isms in rock blasting operations appears to be a prom-
challenge. ising basis for an approach to blastability assessment
The development of rock blasting has a relatively problems.
short but signi®cant history [8] and the advent of rock The RES approach contains a very useful procedure
mass classi®cation systems [37, 38] has made signi®cant for devising a rock mass classi®cation scheme for any
impacts on the assessment of rock mass quality. rock engineering project. In a rock mass classi®cation
However, a generic methodology for the appraisal of scheme, a single parameter is required to comprehen-
the blastability of a rock mass encountering a standard sively characterise the quality of any rock mass for a
blasting operation remains lacking. One of the reasons given engineering project that is to take place within
has been both the diversity of factors in¯uencing the the rock mass. According to the RES approach, all
blastability of rock and the complexity of the associ- possible rock mass classi®cation schemes can be rep-
ated representation of all the in¯uences of the various resented by a function of the leading diagonal par-
factors and the interactive mechanisms between them. ameter values of an interaction matrix. The selection
Another has been the temptation to include controlla- of the parameters and the de®nition of the weighting
ble factors relating to blast design within the scope of of each parameter in a classi®cation system can be
the term blastability [10, 17] and this has led to a more made through the coding of the interaction matrix fol-
confused approach than one which retains the term for lowing a rational procedure. This coding is crucial to
intrinsic rock mass properties. the applicability of the equation in the classi®cation
Several questions then remain. Of the many di€erent scheme. The RES approach has been applied to a
parameters thought to be important, how can dupli- number of rock engineering ®elds, for example, the
cation of similar parameters and the over-in¯uence of assessment of stability of underground excavations [39],
minor parameters be avoided? How would all import- hazard and risk assessment of rockfall [40] and rock
ant factors which cannot be easily expressed with mass characterisation for indicating natural slope
objective measurements be taken into account? How instability [41]. The approach forms one key stage in
can the interactions between the individual contribu- establishing the blastability system.
tory factors be described and presented? It is quite dif-
®cult to answer these without recourse to novel 4.1.2. The interaction matrix and its coding
methodologies. The new methodologies need to pro- In the RES approach to rock engineering, the inter-
vide a basic analytical tool and a presentational tech- action matrix device [18] is both the basic analytical
nique for characterising all relevant factors and tool and a presentational technique for characterising
interaction mechanisms and then to tailor the quanti- the important parameters and the interaction mechan-
tative use of parameters to tackle the complexity of isms in a rock engineering system. In the interaction
blastability assessment. Rock engineering systems matrix for a rock engineering system (e.g. a blastability
methodology [18] was introduced in response to the system), all factors (or parameters) in¯uencing the sys-
need for an ``all-encompassing'' procedural technique tem are arranged along the leading diagonal of the
to approach increasingly complex rock engineering matrix, called the diagonal terms. The in¯uence of
problems, and this appears to be one promising avenue each individual factor (or parameter) on any other fac-
for the development of a blastability assessment sys- tor (or parameter) is accounted for at the correspond-
tem. It is dicult to imagine the practical data acqui- ing o€-diagonal position, and these are named the o€-
sition of suciently complete data set from which diagonal terms. The o€-diagonal terms, are assigned
conventional, e.g. multivariate statistical approaches values which describe the degree of the in¯uence of
46 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

Fig. 2. Illustration of the interaction matrix in RES (from Hudson [18]) (a) Interaction matrix of two factors, (b) general illustration of the cod-
ing of interaction matrix and the set-up of the cause and e€ect coordinates.

one factor (or parameter) on the other factor (or par- and the sum of a column is the ``e€ect'' value, desig-
ameter). Assigning these values is called coding the nated as coordinates (C, E) for a particular factor. C
matrix. A problem containing only two factors is the represents the way in which Pi a€ects the rest of the
simplest example of the interaction matrix, as shown system and E represents the e€ect that the rest of the
in Fig. 2(a). system has on Pi. The coordinate values for each fac-
A general illustration of the coding of interaction tor can be plotted in cause and e€ect space, forming a
matrix is shown in Fig. 2(b). The row passing through so-called C±E Plot [18]. After obtaining the C±E plot
Pi represents the in¯uence of Pi on all the other factors for a system, an equation de®ning a classi®cation
in the system, while the column through Pi represents index that takes into account key contribution factors
the in¯uence of the other factors, or the rest of the sys- can be developed. These stages are shown in Figs. 3
tem, on the Pi. Several procedures have been proposed and 4.
for numerically coding this matrix, for example, the 0± In principal, there is no limit to the number of fac-
1 binary and the expert semiquantitative (ESQ) tors (or parameters) that may be included in an inter-
method [18] and the continuous quantitative coding action matrix, although the number of factors (or
(CQC) method [42]. After coding the matrix by insert- parameters) needed to solve a practical engineering
ing the appropriate values for each cell of the matrix, problem are ®nite. A problem which includes n factors
the sum of each row and of each column can be calcu- (or parameters) will have an interaction matrix with n
lated. The sum of a row is termed the ``cause'' value rows and n columns, as shown in Fig. 2(b).

Fig. 3. Developing a rock engineering classi®cation system by means of the interaction matrix (from Hudson [18]). (a) forming the ordered histo-
gram, (b) formulating the rock classi®cation index.
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 47

Fig. 4. Illustration of the interaction matrix coding results. (a) Coding values, (b) the C±E plot and (c) the ordered histogram (see Section 4.2).

4.2. Formulating the blastability assessment parameter'' column in Table 2). The factors and their
depicting parameters are listed below:
We now return to the development of a blastability
system. Firstly, we select the factors in¯uencing the P1 Strength, represented by uniaxial compression
blastability. Identi®cation of relevant factors can be strength (UCS) of intact rock or point-load
obtained from an extensive review of literature on strength index (PLI);
blasting (e.g. see references listed in Section 4.1.2) com- P2 Resistance to fracturing, represented by the uni-
bined with the authors' experience and judgement. The axial tensile strength (UTS);
following 12 factors (see the ``factors a€ecting blast- P3 Sturdiness, represented by density of rock (r);
ability'' column in Table 2) were chosen as the basic P4 Elasticity, represented by static or dynamic mod-
ones to be considered in establishing a blastability ulus of rock (E);
classi®cation system for a general site, i.e. these 12 fac- P5 Resistance of rock to dynamic loading, rep-
tors were chosen as the diagonal terms in the inter- resented by P-wave velocity (Vp);
action matrix used to establish the blastability system. P6 Hardness of rock, represented by Schmidt ham-
The matrix might be coded by means of subjective jud- mer rebound value (SHV);
gement and experience or objective measurements, or P7 Deformability, represented by Poisson's ratio (m);
both. However, relating to each of these 12 factors, P8 Resistance of rock to breaking, represented by
one (or two) measurable parameter(s) that can, to fracture toughness of rock (KIc);
some extent, depict the factor's in¯uence at a given P9 In-situ block size of rock mass, represented by
site, has been used as the diagonal term to represent mean in-situ block size (MIBS) or principal mean
this factor in the interaction matrix (see the ``depicting spacing (PMS);
48 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

Table 2
Suggested quantitative indications for the classi®cation of the blastability of a rock mass associated with individual factor [22]

Description of ease of blasting Blastability class

Very easy Easy Moderate Dicult Very dicult


Pi Factors a€ecting blastability Depicting parameter 1 2 3 4 5

P1 Strength uniaxial compressive strength <25 25±60 60±100 100±180 >180


(UCS) (MPa)
point-load strength index (MPa) <1 1±2.5 2.5±4 4±9 >9
P2 Resistance to fracturing uniaxial tensile strength <1.5 1.5±3 3±6 6±12 >12
(UTS) (MPa)
P3 Sturdiness of rock density, r (t/m3) <2.0 2.0±2.4 2.4±2.75 2.75±3.0 >3.0
P4 Elasticity of rock E(GPa) <25 25±50 50±100 100±150 >150
P5 Resistance to dynamic P-wave velocity (km/s) <1.5 1.5±2.5 2.5±3.0 3.0±4.0 4.0
loading
P6 Hardness of rock Schmidt hardness value <15 15±30 30±40 40±50 >50
P7 Deformability Poisson's ratio >0.35 0.3±0.35 0.25±0.30 0.25±0.20 <0.20
P8 Resistance to breaking fracture toughness of rock (MPa.m1/2) <0.5 0.5±1.5 1.5±2.5 2.5±3.5 >3.5
P9 In-situ block sizes mean IBSD (m) <0.25 0.25±0.75 0.75±1.5 1.5±2.5 >2.50
mean spacing (m) <0.1 0.1±0.5 0.5±1.5 1.5±2.5 2.5.3
P10 Fragility of rock mass fractal dimension of <1.50 1.50±2.00 2.00±2.50 2.50±2.75 >2.75
in-situ rock mass, D
P11 Integrity of rock mass ratio of P-wave in ®eld to that in lab, <0.35 0.35±0.55 0.55±0.75 0.75±0.9 >0.90
Rv
RQD (%) <40 40±60 60±75 75±90 >90
P12 Discontinuity plane's cohesion C (MPa) <0.05 0.05±0.15 0.15±0.25 0.25±0.50 >0.50
strength
friction angle j (8) <7.5 7.5±15 15±20 20±30 >30

P10 Fragility of rock mass, represented by fractal factors to be included, it may not be possible to ®nd
dimension of in-situ block sizes (D); measurable parameters that fully quantify each factor.
P11 Integrity of rock mass, represented by the wave However, to simplify the system to manageable and
velocity ratio, Rv (the ratio of P-wave velocity in relatively easily obtained ®eld parameters, the set of
the ®eld to that for the laboratory-size specimen), depicting parameters above were chosen.
or by RQD; Using either the ESQ or the CQC coding method,
P12 Discontinuity plane's strength, represented by the coding values, the C±E plot and the ordered histo-
cohesion, C, or friction angle, f of dominant set gram, all of which re¯ect the interaction intensity for
of discontinuities. each of the factors, can be obtained (see Fig. 4, which
is the result of a case study that is described in the
next section). It is important to bear in mind that the
Naturally, one can include more, such as the discon- coding values would probably vary according to di€er-
tinuity orientation related to the face to be blasted. ent opinions from di€erent researchers and ideally sev-
Bhandari [48] refers to his model experiments in which eral experts' opinions should be involved in the factor
fragmentation processes are shown in photographs to selection and coding process. The next stage after cod-
be di€erent when dominant joint sets take the di€erent ing is to calibrate the parameters with greatest inter-
orthogonal planes with respect to the blast face. Roy action intensity and contribution to the blastability of
and Dhar [49] go further and conclude that best frag- a rock mass based on the geological and geotechnical
mentation occurs when the dominant joint face strikes information.
at between 258 and 658 to the blast face. Making use of the ®ndings of blasting theory and
Quanti®cation and weighting of such an orientation practice, a quantitative list of classes of blastability
e€ect will require further research as the situation is connected to individual factors and their depicting par-
complex and is in¯uenced by the free face and thus the ameters is suggested in Table 2. This table is the basis
detonation sequence. Cunningham's algorithm [4, 5] for both rating the in¯uence of each parameter on the
for the rock factor A provides some general guidance blastability and obtaining the rating value used in
on likely degrees of in¯uence. Eq. (6) below. With reference to Table 2, continuous
Water content in the rock mass, joint aperture and rating charts corresponding to each single factor have
various other parameters could also be signi®cant vari- been created to help for borderline cases and also to
ables to take into account. Having chosen the various remove an impression that abrupt changes in ratings
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 49

This contrasts with Bi which is lower for a rock that is


more dicult to blast. However, both BD and Bi have
the same physical signi®cance. To progress with the
application of the E-B-T model a relationship between
them is required. Unfortunately, a database of blasting
operations with parameters from Table 2 is not avail-
able at present. An examination of the possible range
of the values of both Bi and BD does provide a hint of
the preliminary relationship.
The values of speci®c charge are usually in the range
0.15±0.7 kg/m3, the ratio of Sai/Sab is usually in the
range 2±10 and the range of BD is usually 0.2±0.9,
Fig. 5. The suggested continuous rating chart for mean in-situ block
size of rock mass.
thus the range of Bi is estimated to be from 5 to 60
(m1/2/kWh/t). Combining these ranges with the experi-
ence and results from case studies (referred to in
occur between classes. As an example, the continuous Section 5), an empirical equation relating BD and Bi is
rating chart for mean in-situ block size is illustrated in tentatively suggested [22] as follows (Fig. 6).
Fig. 5.
10
It is possible to include as many factors as might Bi ˆ : …7†
conceivably a€ect the blastability. However, only fac- BD
tors which make major contributions to the blastability
system will be of interest for practical application, as Recalling the implications of BD and Bi, a prelimi-
this can probably give a relatively good approximation nary classi®cation for blastability of rock masses
while reducing the burden of data collection. Based on according to the E-B-T coecient Bi or BD has there-
the associated C±E plot (see Fig. 9.6(b) of [18] for the fore been suggested, as shown in Table 3 [22].
signi®cance of the C±E plot in selecting the ®nal con- The Kuz±Ram equation [4] has been widely used to
tributory factors) and the ordered histogram obtained, predict block sizes of blastpiles for a given blast de-
those factors which contribute to most of the system, sign. One of the main challenges to improve the
say larger than 70% of the a(C + E) total in the method is to de®ne the rock factor, A which is often
ordered histogram, can be selected as the factors to be roughly selected by rules of thumb or by improve-
used in assessing the blastability of the rock mass. ments which use empirical formulae. Based on Lilly's
The assessment of blastability of the rock mass can work [16], Cunningham [5] proposed an algorithm for
then be made according to the following formula [22] calculating the values of A. This algorithm took four
X
n factors into consideration and improved the appli-
BD ˆ Wj R j , …6† cation of the Kuz±Ram equation. As demonstrated
jˆ1 above, BD is developed using a systems approach and
it is based on a more comprehensive range of both
where BD is a designation which collectively quanti®es intact rock properties and discontinuity structures than
the resistance to fragmentation by blasting of a rock A in Cunningham's algorithm. The use of BD in deter-
mass and is hereafter called the blastability desig- mining a value for A would appear to be an improve-
nation. Rj is the rating value of the jth factor obtained ment upon Cunningham's algorithm for use with the
from either Table 2 or preferably the corresponding Kuz±Ram equation. A tentative empirical equation
continuous rating charts, according to values indirectly
derived or measured from site. Wj is the weighting
coecient determined from the jth factor according to
its contribution to the system, which can be calibrated
from the ordered histogram. It is obvious that the
value of BD is in the range 0 to 1, and that the greater
the BD, the more dicult the rock is to blast.

4.3. Relationship between BD and Bi

As discussed above, BD is designed to give a com-


prehensive measure of the blastability of a rock mass.
The value of BD is in the range of 0 and 1, and the
greater is BD, the more dicult the rock is to blast. Fig. 6. Suggested empirical relationship of BD and Bi.
50 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

Table 3
Blastability classi®cation according to the E-B-T coecient Bi and BD

Description of ease of blasting Very easy Easy Moderate Dicult Very dicult

Blastability class 1 2 3 4 5
Bi (m0.5/kWh/t) >40 20±40 13±20 8±13 <8
BD <0.25 0.25±0.50 0.50±0.70 0.70±0.85 >0.85

relating BD and A is therefore suggested as follows For the highway improvement, a new route nearly
600 m long was to be created in a deep cutting to be
A ˆ 13  BD: …8†
excavated by blasting. The road cutting was divided
into berms, the depth of a berm was generally 4±6 m.
The examination of Eq. (8) will be described in At the time of the authors' site investigation, the cut-
Section 5. ting had been excavated down to the second berm and
there were one or two deeper berms to be cut. The
rock types at the site were seen to include siltstones,
sandstones, tutes, tu€s and limestones.
5. Blastability: a case study
5.2. Site investigation and data collection
The blastability assessment system developed above
has been applied to a case study that assesses the blast- The available data from the previous geological in-
ability of the rock mass at a highway improvement vestigations, possibly restricted by the limited ex-
cutting site in North Wales (hereafter referred to as posures, were inadequate to provide a satisfactory
the G cutting site). This application served as one of explanation for why a high percentage of blasting
the ®rst trials of the blastability system and it was failed to obtain satisfactory fragmentation. In order to
found to be useful, while investigating the reasons for provide an up-to-date assessment of blastability of
blasting problems encountered at the site. rock materials at the site, the authors undertook
further geological data acquisition. The investigation
5.1. Background of the case study carried out involved mapping discontinuities on var-
ious rock cuttings, taking photos of blasting results im-
The main concern in the case study was to identify mediately after blasting, performing on-site point load
the reasons for the high percentage of blasts which tests and Schmidt Hammer tests and collecting other
produced unsatisfactory fragmentation (i.e. excess associated geological and blast design data. A sketch
oversize). plan for the investigation, together with the positions

Fig. 7. A sketch plan for the geological investigation at the G cutting site showing locations of scanline mapping and intact rock samples.
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 51

Table 4
Blastability assessment of rock masses at the G cutting site

Parameter Weight, Wi Blastability assessment

description unit S1 S2 S3

No. Pi value rating Wi*Ri value rating Wi*Ri value rating Wi*Ri

1 Is(50) MPa 0.1475 4.42 0.65 0.096 6.51 0.8 0.118


UCS MPa 120 0.680 0.100
2 UTS MPa 0.1344 6.82 0.670 0.090 5.53 0.580 0.078 8.14 0.700 0.094
4 E GPa 0.1273 48.80 0.395 0.050 45.13 0.380 0.048 52.39 0.410 0.052
3 r t/m3 0.1249 2.710 0.675 0.084 2.704 0.680 0.085 2.715 0.720 0.09
6 SHV 0.1225 43.6 0.710 0.087 41.2 0.660 0.081 46 0.750 0.092
V
5 p m/s 0.1208 4901 0.920 0.111 4784 0.900 0.109 4996 0.930 0.112
10 D 0.1131 1.486 0.440 0.05 1.397 0.420 0.048 1.848 0.590 0.067
9 MIBS m 0.1095 2.18 0.830 0.091 3.10 0.935 0.102 2.80 0.915 0.100
Blastability designation 0.664 0.647 0.725

No. Pi Parameter Weight, Wi Blastability assessment

S4 S5 S6

description unit value rating Wi*Ri value rating Wi*Ri value rating Wi*Ri

1 Is(50) MPa 0.1475 4.45 0.65 0.096 5.05 0.69 0.102


UCS MPa 135 0.730 0.108
2 UTS MPa 0.1344 5.563 0.590 0.079 7.67 0.710 0.095 6.31 0.680 0.091
4 E GPa 0.1273 50.07 0.400 0.051 48.47 0.400 0.050 42.9 0.330 0.042
3 r t/m3 0.1249 2.70 0.675 0.084 2.63 0.610 0.076 2.70 0.680 0.084
6 SHV 0.1225 44.85 0.725 0.089 46.07 0.740 0.091 39.7 0.550 0.067
V
5 p m/s 0.1208 4785 0.900 0.109 4908 0.920 0.111 4852 0.910 0.110
10 D 0.1131 2.113 0.690 0.078 1.499 0.450 0.051 1.194 0.370 0.042
9 MIBS m 0.1095 1.31 0.690 0.076 1.78 0.790 0.087 2.22 0.830 0.091
Blastability Designation 0.662 0.669 0.631

of the scanline mapping, the point load tests and the samples, tests and experience. Due to a lack of com-
Schmidt Hammer tests is illustrated in Fig. 7. plete sets of test results, a number of empirical for-
mulae based on published correlation studies [43±46]
5.3. Blastability assessment have been used to derive missing parameters. The
results obtained from the site investigation showed the
With help from the resident engineers and consult- varying degrees of both mechanical properties and dis-
ants involved in the cutting, the interaction matrix of continuity structures from one place to another. This
the blastability system was coded using the CQC is re¯ected by the blastability assessment results, as
approach [42], and the coding results were obtained as shown in Table 4.
illustrated in Fig. 4. The data taken from the authors' site investigation
It is seen from Fig. 4 that the range in parameter in- and sampling based on six ®eld locations along the
teraction intensity is quite wide (refer to Fig. 9.6(b) cutting, when subjected to the blastability analysis,
in [18]). Thus, only those factors contributing to a yielded values of the blastability designation from
total of 72.5% of the a(C + E) in the ordered histo- 0.631 to 0.725 (i.e. Bi from 15.85 to 13.75 m1/2/kWh/t),
gram, that is, the eight parameters, P1, P2, P3, P4, P5, as shown in Table 4. This indicated that the rock
P6, P9, P10 have been chosen as the main contributory masses in the highway cutting area belong basically to
factors of the blastability of the rock masses at the the border range between class 4 (dicult blasting)
site. The corresponding weights of the eight factors and class 3 (moderate blasting), see Table 3. Applying
were derived using the method illustrated in Fig. 4, the description terms to the BD results, this means
and they are listed in Table 4 (column Wi). Having that the rock masses are, in general, dicult or moder-
completed the ®rst stage, which is concerned with the ate to blast.
matrix coding and thus the parameter weighting, the The blastability of the rock mass at this site was
second stage is to obtain the actual results for each also assessed at three out of the six locations utilising
parameter (column value in Table 4) using ®eld only the previous geological information that was
52 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

Fig. 8. Comparison of BBSDs from the E-B-T model with alternative prediction methods. (a) 31/07/95 blast, (b) 03/08/95 blast.

available. The values of BD obtained were from 0.498 Ram model which can take advantage of the new
to 0.569, indicating that the values of BD belong to IBSD estimation procedure [6, 20], (iv) estimation
class 2 or the border range between class 2 and class 3 using the new E-B-T model and blastability assessment
(see Table 3). This suggests that the new ®eld test data and (v) an estimation of the BBSD primarily based on
has revealed that the rock mass is generally more di- the Kuz±Ram model but with a correction such that
cult to blast than might have been expected from the the rock factor in the Kuz±Ram model, A, is deter-
initial site investigation. mined by Eq. (8) (such a model is called the corrected
Kuz±Ram model).
5.4. Estimation of BBSD based on blastability A comparison of the BBSD results of two blasts,
assessment one carried out on July 31 and another on August 3,
1995, which considers the di€erent estimation tech-
It is interesting to consider that the blastability niques mentioned above, is illustrated in Fig. 8. The
assessment results obtained above are re¯ected in the blast pattern data, the borehole parameters and the
actual blasting results which indicated by the BBSDs explosive details are summarised in Table 5. From
of the blastpiles obtained. To examine the BBSD at Fig. 8, the following important observations can be
the site, the following approaches will now be com- made.
pared: (i) direct assessment of the BBSD by the photo- The BBSD directly assessed using the photo-scanline
scanline method devised by the authors [47], (ii) esti- method appears to lie near the average of the predic-
mation using the previously published Kuz±Ram blast tions from the other four techniques. The BBSD pre-
design model [4] with no IBSD information needed, dictions from the Kuz±Ram and Bond±Ram models
(iii) estimation using the previously published Bond± form the far upper and far lower boundaries while

Table 5
Parameters associated with the predictions of BBSD

Blasting pattern parameters Rock mass parameters

Date of blast Date of blast


31/07/95 03/08/95 Blasting site 31/07/95 03/08/95

Q (kg) 1430 605 Is(50) (MPa) 5.05 4.45


q (kg/m3) 0.65 0.62 UCS (MPa) 111.10 97.90
Bench height (m) 7 7 UTS (MPa) 6.31 5.56
Hole depth (m) 8 8 E (GPa) 46.10 50.10
Subdrill (m) 1 1 SHV 39.70 44.85
Hole No.s 55 22 mean IBSD (m) 2.20 1.30
Hole diameter (mm) 105 105 r (t/m3) 2.70 2.70
Explosive PG800/900 PG800/900 Vi50 (m3) 6.57 1.33
Burden (m) 2.5 2.5 Vi63.2 (m3) 13.78 2.61
Spacing (m) 2.5 2.5 nv 0.938 0.554
Bottom charge (m) 1 1 Si50 (m) 2.21 1.30
Column charge (m) 3 3 Si63.2 (m) 2.83 1.63
Stemming (m) 3 3 ns 2.813 1.663
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 53

that from either the newly developed E-B-T model or the assessment system, it is possible to de®ne the value
the corrected Kuz±Ram model based on the blastabil- of Bi, and thereby use the E-B-T model as a blast de-
ity assessment is approximately in the middle of the sign tool.
range formed by BBSDs from the Kuz±Ram and the A sample of available data sets from the literature
Bond±Ram models. Also, the BBSD from the E-B-T suggests that the E-B-T model would generally give an
model and the corrected Kuz±Ram model are close to improvement in describing the energy±size reduction in
the BBSD assessed using the photo-scanline technique blasting compared with the Bond±Ram model. The
for the 03/08/95 blast. case study presented also suggests that the proposed
blastability assessment system and the associated E-B-
5.5. Discussion of case history T model represents a possible improvement on existing
models. The calculation of the rock factor A in the
It seems reasonable to assume that early blasting op- Kuz±Ram model based on the blastability assessment
erations at the site would have begun by taking instead of Cunningham's A also appears to give an
account of little more than geological data from the in- improvement in the accuracy of the Kuz±Ram
itial site investigation. But, it has been shown that the equation when applied to the case study data.
blastability designation based on rock mass data However the parameter set developed in the paper to
revealed during excavation is greater than that likely assesses blastability has yet to include the in¯uence of
to be estimated from the previously available data set. the dominant joint set orientation with respect to the
Furthermore, the variability of blastability might not free face.
have been previously recognised. Not taking account A well-developed model or system should be
of the variability in blastability probably contributed exposed to a variety of case study examinations with
to unsatisfactory blasting results in various places. A di€erent geological conditions in order to achieve su-
possible opportunity to optimise is indicated by the cient con®dence for its use. This could be most e-
di€erence between BD of 0.631 (site S6) and 0.725 ciently implemented by the setting up of a database
(site S3) (see Table 4) which would represent a clear with the full record of di€erent in-situ geological con-
increase in speci®c charge required. ditions, blast patterns, explosive energy inputs and
directly assessed BBSD results. A series of model-scale
or full-scale trial blasts would be ideal for further vali-
6. Concluding remarks dation of the E-B-T model and the associated blast-
ability assessment. The relationship proposed between
The E-B-T model proposed for characterising blast- the E-B-T coecient and the blastability designation,
ability Ð the ease with which a rock mass can be frag- Eq. (7) has a comparatively poor level of con®dence
mented by blasting Ð has been outlined. The E-B-T and therefore further calibration is to be rec-
coecient Bi and E-B-T model have been proposed to ommended.
account for the e€ects that rock masses with di€erent The blastability assessment system presented
blastability will have. includes a signi®cant amount of subjective criteria
Recognising that blastability is a composite intrinsic through the matrix coding procedure. Systems
property and that conventional approaches are not approaches such as the rock engineering systems and
well suited to characterising the blastability, a method- the grey systems [42] have been exploited to reduce the
ology for the assessment of blastability of rock masses ojectivity. Further investigations into how to more
using the rock engineering systems approach has been subjectively represent parameters of the blastability
developed in this paper. The methodology has system- system and how to more accurately code the inter-
atically taken into account twelve factors which a action mechanisms in the matrix are necessary for
review suggested would give a reasonably comprehen- improvement in the blastability assessment.
sive set of factors that in¯uence the blastability of a In this paper the energy input for the model has
rock mass. The contribution of each of these factors to only been related to uncontrollable factors governed
the blastability of the rock mass is identi®ed using in- by in-situ geological conditions and the term blastabil-
teraction matrix analysis, which is implemented by de- ity has been deliberately restricted to quantify this
riving a weighting for each factor. Combining the intrinsic resistance of the rock mass. In the companion
results from the interaction matrix analysis and the paper [23], it is brie¯y suggested that the controllable
ratings proposed in this research, the blastability of factors such as burden, spacing, delays, decoupling,
the rock mass may be represented quantitatively using etc. could be introduced into the E-B-T model through
the blastability designation, BD. A preliminary classi®- a composite coecient fc that regulates the e€ective-
cation for blastability of rock masses according to the ness of the input energy of explosives. The utility of
E-B-T coecient or the blastability designation has such an E-B-T model with or without fc remains some-
therefore been suggested. Using the BD derived from what speculative.
54 J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55

References [23] Lu P, Latham, J-P. A model for the transition of block sizes
during rock blasting. Int. J. Blast. Fragment., in press.
[1] Belland JM. Structure as a control in rock fragmentation. CIM [24] Charles RJ. Energy±size reduction relations in comminution.
1966;59:323±8. Trans. AIME 1957;208:80±8.
[2] Hagan TN, Just GD. Rock breakage by explosives: theory, [25] Schuhmann R, Jr. Energy input and size distribution in commi-
nution. Trans. Am. 1st Min. Metall. Petrol. Engrs. 1960;217:22±
practice and optimisation. In: Proc. 3rd Congr. ISRM. National
5.
Academy Sciences, Washington, DC, 1974;1(B):1349±1358.
[26] Nagahama H, Yoshii K. Fractal dimension and fracture of
[3] Yang Z-G, Rustan A. The in¯uence from primary structure on
brittle rocks. Int. J. Rock Mech. Min. Sci. Geomech. Abstr.
fragmentation. In: Proc. 1st Int. Symp. Rock Fragmentation by
1993;30(2):173±5.
Blasting, Lulea, Sweden, 1983;581±603.
[27] Wang H. Predictions of in-situ and blastpile block size distri-
[4] Cunningham C. The Kuz±Ram model for prediction of frag-
butions of rock masses, with special reference to coastal require-
mentation from blasting. In: Proc. 1st Int. Symp. Rock
ments. Ph.D. thesis, Queen Mary and West®eld College,
Fragmentation by Blasting. Lulea, Sweden, 1983:439±453.
London University, 1992.
[5] Cunningham C. Fragmentation estimation and the Kuz±Ram
[28] Aler J, Du Mouza J, Arnould M. Measurement of the fragmen-
model: four years on. In: Proc. 2nd Int. Symp. Rock
tation eciency of rock mass blasting and its mining appli-
Fragmentation by Blasting. Colorado, U.S.A., 1987:475±487.
cations. Int. J. Rock Mech. Min. Geomech. Abstr.
[6] Da Gama CD. Use of comminution theory to predict fragmen-
1996;33(2):125±39.
tation of jointed rock mass subjected to blasting. In: Proc. 1st
[29] Aler J, Du Mouza J, Arnould M. Evaluation of blast fragmen-
Int. Symp. Rock Fragmentation by Blasting. Lulea, Sweden,
tation eciency and its prediction by multivariate analysis pro-
1983:563±579.
cedure. Int. J. Rock Mech. Min. Geomech. Abstr.
[7] Clarke GB. Principles of Rock Fragmentation. John Wiley,
1996;33(2):189±96.
New York, 1987.
[30] Franklin JA, Broch E, Walton G. Logging the mechanical char-
[8] Atlas Powder Company. Explosives and Rock blasting. Atlas
acter of rock. Trans. IMM 1971;A:1±9.
Powder Company, Texas, 1987.
[31] Fourney WL, Barker DB, Holloway DC. Fragmentation in
[9] Wang H, Latham J-P, Poole AB. Blasting design for armour- jointed rock materials. In: Proc. 1st Int. Symp. Rock
stone production. Quarry Management, 1991, part I (July), 17± Fragmentation by Blasting, Lulea, Sweden, 1983;505±531.
21; part II (Aug.), 19±22. [32] Rustan A, Vutukuri VS, Naarttijarvi T. The in¯uence from
[10] Scott A, Chitombo G, Kleine T. The challenge of the prediction speci®c charge, geometric scale and physical properties of hom-
and control of fragmentation in mining. In: Rossmanith editor, ogenous rock on fragmentation. In: Proc. Int. Symp. Rock
Proc. of the 4th Int. Symp. -FRAGBLAST-4. Balkema, Fragmentation by blasting, Lulea, Sweden, 1983;115±142.
Rotterdam, 1993:507±517. [33] Rorke AJ. A scienti®c approach to blast design. Quarry
[11] Singh DP, Sarma KS. In¯uence of joints on rock blasting: a Management. 1988;October:0±0.
model scale study. In: Proc. 1st Int. Symp. Rock Fragmentation [34] Stagg MS, Otterness RE, Siskind DE. E€ects of blasting prac-
by Blasting, Lulea, Sweden, 1983:533±554. tice on fragmentation. In: Proc. 33rd U.S. Symp. Rock Mech.,
[12] Jimeno CL, Jimeno EL, Carcedo FJA. Drilling and blasting of Balkema, Rotterdam, 1992;313±322.
rocks. Balkema, Rotterdam, 1995. [35] Lizotte YC, Scoble MJ. Geological control over blast fragmen-
[13] Bond FC, Whittney BB. The work index in Blasting Quarterly tation. CIM Bull. 1994;87(983):57±71.
of the Colorado School of Mines 1959;54(3):77±82. [36] Matheson GD. Aspects of highway rock engineering in the
[14] Just GD. The application of size distribution equations to rock U.K. Engineering Geology of Construction, ed. M. Eddleston et
breakage by explosives. Rock Mech. 1973;5(3):151±62. al. Geol. Soc., London, 1995;169±187.
[15] Kuznetsov VM. The mean diameter of the fragments formed by [37] Bieniawski ZT. Engineering classi®cation of jointed rock mass.
blasting rock. Soviet Mining Sci. 1973;9(2):144±8. Trans. S. Afr. ICE 1973;15:335±44.
[16] Lilly PA. An empirical method of assessing rock mass blastabil- [38] Barton N, Lien R, Lunde J. Engineering classi®cation of rock
ity. In: Davidson editor. Large Open Pit Mining Conference, masses for the design of tunnel support. Rock Mech.
AusIMM, Victoria, 1986;89±92. 1974;6(4):189±236.
[17] Franklin JA, Maerz, NH. Empirical design and rock mass [39] Lu P, Hudson JA. A fuzzy evaluation approach to the stability
characterisation. In: Franklin, Katsabanis, editors. of underground excavations. ISRM Symposium: EUROCK'93,
Measurement of Blast Fragmentation, Balkema, Rotterdam, ed. L Ribeiro e Sousa and NF Grossmann. Balkema,
1996. Rotterdam, 1993;615±622.
[18] Hudson JA. Rock Systems Engineering: Theory and Practice. [40] Cancelli I, Crosta G. Hazard and risk assessment in rockfall
Ellis Horwood, Chichester, 1992. prone areas. In: Skipp B, editor. Risk and Reliability in Ground
[19] JKMRC. Advanced blasting technology. AMIRA P93D (1987± Engineering, Thomas Telford, London, 1993, pp. 177±190.
1990), ®nal report, JKMRC. University of Queensland, [41] Mazzoccola DF, Hudson JA. A comprehensive method of rock
Australia, 1991. mass characterisation for indicating natural slope instability.
[20] Wang H, Latham J-P, Matheson GD. Design of fragmentation QJEG 1996;29:37±56.
blasting in surface excavation. In: Hudson, editor. Rock [42] LuP, Latham J-P. A continuous quantitative coding approach
Characterisation, Thomas Telford, London, 1992;233±238. to the interaction matrix in rock engineering systems based on
[21] Lu P, Latham J-P. In-situ block size distribution prediction grey systems approaches. In: Proc. 7th Int. Cong. of IAEG.
with special reference to discontinuities with fractal spacing dis- Balkema, Rotterdam, 1994;4761±4770.
tribution. In: Proc. ISRM: EUROCK'96. Balkema, Rotterdam, [43] ISRM. Suggested method for determining point load strength.
1996;311±318. Int. J. Rock Mech. Min. Geomech. Abstr. 1985;22(2):53±60.
[22] Lu P. The characterisation and analysis of in-situ and blasted [44] Karpuz C, Pasamehmetoglu G, Bozdag T, MuÈftuÈoglu Y.
block size distribution and the blastability of rock masses. Rippability assessment in surface coal mining. In: Singhal RK,
Ph.D. thesis, Queen Mary and West®eld College, London Vavra M, editors. Mine Planning and Equipment Selection,
University, 1997. Balkema, Rotterdam, 1990;315±322.
J.-P. Latham, P. Lu / International Journal of Rock Mechanics and Mining Sciences 36 (1999) 41±55 55

[45] Sachpazis CI. Correlating Schmidt hardness with comprehensive with Ros±Ram and Schuhmann models. In: Proc. Int. Symp.
strength and Young's modulus of carbonate rocks. Bull. IAEG Mining Sci. Tech. Balkema, Rotterdam, 1996;683±688.
1990;42:75±83. [48] Bhandari S. Changes in fragmentation processes with blasting
[46] Xu S, Grasso P, Mahtab A. Use of Schmidt hammer for esti- conditions. In: Mohanty B, editor. Rock Fragmentation by
mating mechanical properties of weak rock. In: Price DG, edi- Blasting, Balkema, Rotterdam, 1996;301±309.
tor. Proc. 6th Int. IAEG Cong., Balkema, Rotterdam, [49] Roy PP, Dhar BB. Fragmentation analysing scale: a new tool
1990;511±519. for rock breakage assessment. In: Mohanty B, editor. Rock
[47] Lu P, Latham J-P, Yin J. Estimation of blasted block size distri- Fragmentation by Blasting (abstract), Balkema, Rotterdam,
bution of a blastpile: combining ``photo-scanline'' technique 1996;448.

Das könnte Ihnen auch gefallen