See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/30389101

A Comparison Between Hoek-Brown and Bieniawski Criteria for Coal and Rocks

Article · January 1998

Source: OAI

CITATIONS READS
3 460

1 author:

Mohammad Farouq Hossaini

UNSW Sydney
43 PUBLICATIONS   680 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Back Analysis using Ant Clony Algorithm View project

Stress analysis in underground excavations View project

All content following this page was uploaded by Mohammad Farouq Hossaini on 25 December 2017.

The user has requested enhancement of the downloaded file.

 University of Wollongong
Research Online
Coal Operators' Conference Faculty of Engineering and Information Sciences

1998

A Comparison Between Hoek-Brown and

Bieniawski Criteria for Coal and Rocks
S. M. Hossaini
Teheran University

Publication Details
This conference paper was originally published as Hossaini, SMF, A comparison between Hoek-Brown and Bieniawski criteria for coal
and rocks, in Aziz, N (ed), Coal 1998: Coal Operators' Conference, University of Wollongong & the Australasian Institute of Mining
and Metallurgy, 1998, 213-223.

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library:
[email protected]
A Comparison Between Hoek-Brown and Bieniawski Criteria
for Coal and Rocks

s M F Hossainil

ABSTRACT

The applicability of Bieniawski and Hoek-Brown empirical strength criteria has been assessedfor different groups of coal
and various types of intact rocks by using a vast number of published triaxial test data from various places. Analysis of
individual data sets revealed that the traditional forms of the criteria do not have a perfect agreement with the data. A
strong negative correlation has been observed between B in Bieniawski's criterion and m in Hoek and Brown's criterion
with uniaxial compressive strength of materials. Both criteria have been modified and empirical relationships have been
introduced for coal as follows:

m = 62.903 -34.213 (log Pc) 0.9772

B = 10.152- 4.709 (Log Pc )

0.8889

Similar relationships have been developed for different rock types as well.

A comparison between the applicability of each of the above approaches with the conventional criteria reveals a very
significant advantage for new approaches and a supremacy for the Bieniawski criterion in all cases particularly in the case
of coal.

The modified Bieniawski criterion fits coal as well as different types of rocks with excellent accuracy.
The modified Hoek-Brown criterion gives a good result for rocks but does not fit coal data quite well. In other words,
Hoek-Brown criterion is not an suitable one for coal.

INTRODUCTION

To estimate the strength of rock and rock mass a failure criterion is required. The theoretical triaxial strength criteria based
on the actual mechanism of fracture do not fit the experimental results properly and to overcome this problem, many
empirical criteria have been formulated for rocks.

Laboratory strength data values are the starting points for estimating the strength of rock and rock mass. If a criterion fails
to fit the laboratory strength data properly, then its applicability to real field cases would certainly be doubtful.
Out of many strength criteria developed so far, none of them has been accepted to be a global formula capable of simply
describing the strength of geomechanical materials in general.

Coal is one of the most important energy resources. As it is widely mined around the world, and receiving attention on the
current mining and energy scene, developing an appropriate strength criterion for use in coal seams to be of considerable
value.

Amongst all available strength criteria proposed for the estimation of the strength of rocks and geomechanical materials, a
very few of them have been extended to coal. One criterion proposed for coal is that suggested by Sheorey, Biswas and
Choubey(1989). Using this criterion in practice requires a very long complicated procedure, as it is understood from the

1 Senior Lecturer, Department of Mining Engineering, Faculty of Engineering, Teheran University

COAL98 Conference Wollongong 18- 20 February 1998 213

original paper. None of the two failure criteria examined in this investi!!.ation{ i.e. Hoek-Brown. 1980a&
Bieniawski 1974 have taken coal into consideration and the a ro riate values for constants are not available for the
~.

STRENGTH CRITERIA
The general fornl of a strength criterion is:
PI = f (P2. P3)
where p 1. P2 and P3 are the principal stresses at failure.

Because the available data indicate that the intermediate principal stress, P2, has very little influence on strength than thc
minor principal stress, p3, all of the criteria used in practice are reduced to the form:
PI = f (P3)
or in its normalised form;
~ ) = f (.§.1.)
Sc Sc

Hoek-Brown's criterion for intact rocks

Bieniawski's criterion for intact rocks

(2)

B = a constant value for each rock type and

p = 0.65 or 0.75 for all rock type

DA T A SELECTED FOR ANAL YSIS

Coal data representing twenty six seams and collieries are from two publications (Hobbs, 1964 and Das and Sheorey,
1986). These data are homogeneous and on specimens of almost the same size. Thus, they are to a reasonable extent free
from the effects of specimen size.

Intact rocks data include various types of geomechanical materials of diverse lithological and mechanical characteristics
from weak over-consolidated clays of unconfined compressive strength of 24 MPa to strong hard granite and granodiorite
of unconfined compressive strength of 427 MPa.

ANALYSING THE APPLICABILITY OF THE CRITERIA FOR INTACT COAL

Analysis of individual data sets revealed that none of the existing criteria shows perfect agreement with experimental
values of coal strength. Although unique values of the constants in both criteria have been determined with good
coefficients of determination for overall data, a wide variation has been noticed in the values of the constants when
individual data sets have been analysed.

COAL98 Conference Wollongong 18.20 February 1998 214

Hoek and Brown's criterion

The value for m has been detennined to be 25.132 for the combination of all the 180 oairs of the data. The coefficient of
determination has been found to be 0.9105. Plot of all data along with regression curve is shown in Fig. 1.
When the data groups are analysed individually, the out comes differ widely from what is achieved by mixing the whole
data of all groups. In general, the best correlation amongst all single constant values assigned to m was due to m = 10
(suggested for mudstone, siltstone, shale and slate by Hoek and Brown). Fig. 2 shows six examples of those cases for
which by applying this criterion the lowest correlation with exact data has been obtained.

0'3/0' c

Analvsis of individual data sets has !!iven a ran!!e of values from 5.3795 to 50.190 for m. Analysis of these values along
with Pc has indicated that there is a significant correlation between them (Fig. 3). The relationship between m and .c has
been found to be as follows:

(3)

Bieniawski's criterion

A plot of ~ versus ~ for all data along with the regression curve is shown in Fig. 1. The values for B and. have been
Sc Sc
determined to be 3.7062 and 0.9225 respectively. The coefficient of determination has been found to be 0.9551.

COAL98 Conference Wollongong 18- 20 February 1998 215

 Fig. 2 -Examples of discrepancy in the traditional Hoek-Brown criterion for coal.

COAL98 Conference Wollongong 18. 20 February 1998

216
The best result amongst all various constant amounts of B was obtained for the case in which p = 0.65 and B = 3.0. Fig.
4 demonstrates six examples of the cases in which this criterion has given the lowest coincidence with real data (for p
:0.65 and B = 3.0).

from 0.4517 to 0.7423.

The best statistical average for p was found to be 0.6. Taking p as 0.6, the values for parameter B have been recalculated
for all the individual data sets. From this analysis, B has been found to be between 2.0663 and 7.7150 and the relationship
between B and Pc has been found to be as follows:

The coefficient of detennination for this regression has been found to be 0.9164 (Fig. 5). Fig. 6 depicts the correlation
between modified Bieniawski criterion with 6 groups of experimental data for which the lowset coefficient of
detennination was observed. Comparison of Figs. 2 and 4 with Fig. 6 reveals the supremacy of the new version of the
Bieniawski criterion.

ANALYSING THE APPLICABILITY OF THE CRITERIA FOR INTACT ROCKS

The same analysis as conducted for coal was carried out for different rock types, namely, limestone, granite, granodiorite,
shale, sandstone,claystone and liparite. Although the data were from various sources with differences in techniques ,size
and shape of specimens, the results indicate that the parameters can not be regarded as constant values. For each particular
rock type there found to be a correlation between B in the Bieniawski criterion and m in the Hoek-Brown criterion with Pc.
p in the Bieniawski criterion takes different values for different types of rocks. Limestone and granite are taken as two
examples.

COAL98 Conference Wollongong 18.20 February 1998 217

 Fig. 4 -Examples of discrepancy in the traditional Bieniawski criterion for coal

COAL98 Conference Wollongong 18.20 February 1998 218

The values of a and the relationships between B and m with Pc for these two cases are as follows:

for limestone;
a =0.76
B = -3.3538+10.8831og(pc)-0.8131 (5)

m = -1.6115 + 62.05 log(pc)-2.7421 (6)

for granite;
a =0.65
B = 3.4452 + 21.6171og(pc)-2.757 (7)

m = -971 + 1055.3Iog(pc)-O.O6 (8)

"',..

a)
5

2
0 10 20 30 40 50 60

O"c

Fig. 5 Plot of B versus Pc in the Bieniawski's criterion for coal(a = 0.6)

COAL98 Conference Wollongong 18- 20 February 1998 219

 Fig. 6 The lowest correlative cases of the modified Bieniawski criterion.

Figs. 7 and 8 give the relationship between B in the Bieniawski criterion and Pc for limestone and granite as 2 examples.

COAL98 Conference Wollongong 18 -20 February 1998 220

 In

O'c

Fig. 7 Plot of B versus Pc for limestone and marble in the Bieniawski criterionl

01

4
0 100 200 300 400 500

Fig. 8 Plot of B versus Pc for granite and granodiorite in the Bieniawski criterionJ

2 (X = 0.76
3 (X = 0.65

COAL98 Conference Wollongong 18- 20 February 1998 221

 SUMMARY AND CONCLUSIONS

• Although eventual modifications to the selected criteria for intact rocks requires more investigations in which
more proper data groups must be analysed within any rock type, the fIrst conclusion coming out from the
assessments done in this paper implies that treating criteria parameters with constant values would result in
considerable inaccuracy, even for intact materials.

• Tables 1 and 2 summarise the statistical analysis of applying the two modified criteria to the laboratory data of
coal and intact rocks(~ is the coefficient of determination). As shown in these tables, a significant accuracy is
obtained by applying the modified Bieniawski criterion for both coal and intact rocks. This criterion, therefore,
satisfies all the requirements for an desirable empirical strength criterion provided that it is amended with the
modifications suggested in this investigation.

The modified Hoek and Brown criterion gives good level of accuracy for rocks but is not a suitable criterion for coal as
shows low correlation with the coal data (Table 1).

Table 1 Summary of comparison of 2 modified criterion for coal.

Range-of r"- with real data

criterion cases 2 cases 2 cases 2

with r with r ~ with r
?:0.95 0.90% ~0.85
modified Bieniawski 69% 92% 96%

modified Hoek-Brown 38% 58% 65%

Table 2 Summary of comparison of 2 modified criterion for intact rocks

Range of r~ with real data

critenon cases 2 cases 2 cases 2

with r with r ~ with r

?; 0.95 0.90% ~0.85

modified Bieniawski 81% 94% 98%

modified Hoek-Brown 72% 91% 95%

• An estimate of the triaxial strength can be made by meansof the Bieniawski criterion with a variable B dependent
upon Pc and a certain constant p for each particular material. The only parameter required for this criterion is the
unconfined compressive strength which can be determined simply.

• A strength criterion must be capable to deal with different conditions of a certain type of rock having different
properties. Such a criterion mayor may not provide the best estimation for a large number of mixed data from
various collieries and seams around the world.

• In practice, a design engineer is faced with a certain type of rock with its particular properties. The characteristics
of any rock type may change from place to place or even from one part to another part of the same seam or block.
A criterion must be flexible enough to fit the various conditions of rock properties.

COAL98 Conference Wollongong 18 .20 February 1998

222
 ACKNOWLEDGMENT
The author wishes to thank Teheran University for supporting his further research on the failure criteria. Because the
author fIrst started his work on this topic as a part of his study at the University of New South Wales, he wishes to thank
Dr V. S. Vutukuri for his supervision during the author's PhD study.

REFEREN CES

Bieniawski. Z T, 1974. Estimating the strength of rock materials. J. S. African Institute of Mining and Metallurgy, Vol. 74,
No.8. pp. 312 -320.

Das, M N and Sheorey, p R, 1986. Triaxial strength behaviour of some Indian coals. Journal of Mines, Metals & Fuels,
March, pp. 118- 122.

Hobbs, D W, 1964. The strength and the stress-strain characteristics of coal in triaxial compression. J. Geol. No.72
pp.214-231.

Hoek, E and Brown, E T, 1980a. Underground excavations in rock. London. Instn. Min. Metall.

Hoek, E, and Brown, E T, 1980b.Empirical StrengthCriterion for Rock Masses,I. of the GeotechnicalEngng. Division,
ASCE,Vol. 106,No. GT9, Paper 15715,sep,pp. 1013- 1035.

Hossaini, S M F,1994. Some aspects of the strength characteristics of intact and jointed rocks. PhD thesis, University of
New South Wales.

Sheorey, p R, Biswas, A K and Choubey, V D, 1989. An empirical failure criterion for rocks and jointed rock masses.
Engng. Geology, Vol. 26, pp. 141-159.

Vutukuri, V S and Hossaini, S M F 1992a. Assessment of applicability of four empirical strength criteria for intact coal.
Proceedings.6th Australia -New Zealand Conference. Geomchanics (Geomechanical risk identification, evaluation
and solutions), pp. 280-285, New Zealand.

Vutukuri, V S and Hossaini,S M F , 1992b.Assessmentof applicability of strengthcriteria for rock and rock massto coal
pillars. Proceedings.11th 1ntemational.conference.on Ground control in Mining (eds.N.1.Aziz and S. S. Peng). PP.
1-8, Wollongong.

COAL98 Conference Wollongong 18 -20 February 1998 223

View publication stats