S.•/.

/,,AD/A207 597

(ROD) AFTER
ROCK QUALITY DESIGNATION
TWENTY YEARS

Don U. Deere, Consultant

Gainesville, FL

Feb 89

•• ______--------__--____-

U.S. DEPARTMENT OF COMMERCE

National Technical Information Service
I,• '. -8.
9- !:OF':1G
1";,
 CONTRACT REPORT GL-89-1

ROCK QUALITY DESIGNATION (RQD)

AFTER TWENTY YEARS
by

Don U. Deere
Consultant
C Gainesville, Florida 32608

and
I Don W. Deere
0 Geotechnical Engineer
Rocky Mountain Consultants, Inc.
Longmont, Colorado 80501
1

4
0

£
0
U

February 1989
Final Report

Approved For Public Release; Distribution Unlimited

REPRODUCED BY
U.S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL INFORMATION SERVICE
SPRINGFIELD, VA. 22161

DEPARTMENT OF THE ARMY

Prepared for
US Army Corps of Engineers
Washington, DC 20314-1000
Under Contract No. DACW39-86-M-4273
Monitored by Geotechnical Laboratory

US Army Engineer Waterways Experiment Station

PO Box 631, Vicksburg, Mississippi 39181-0631
)
 IINflI ASSTFTFD
SECURITY CLASSIFICATION OF THIS PAGE
I Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0?88

lb RESTRICTIVE MARKINGS
Ia. REPORT SECURITY CLASSIFICATION
Unclassified
3 DISTRIBUTION /AVAILABILITY OF REPORT
2a. SECURITY CLASSIFICATION AUTHORIT'Y
Approved for public release;
2b. DECLASSIFICATION/DOWNGRADING SCHEDULE distribution unlimited.
S MONITORING ORGANIZATION REPORT NUMBER(S)
4. PERFORMING ORGANIZATION REPORT NUMBER(S)

Contract Report GL-89-1

6a. NAME OF PERFORMING ORGANIZATION 6o. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION
(if applicable) USAEWES
Geotechnical Laboratory
6c. ADDRESS (City, State, and ZIP Code) lb. ADDRESS(City, State, and ZIP Code)

P0 Box 631
(See reverse) Vicksburg, MS 39181-0631
NAME OF FUNDING/SPONSORING Sb. (if
OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
Ba. ORGANIZATION app~licable)

US Army Corps of Engineers DACW39-86-M-4273

10 SOURCE OF FUNDING NUMBERS
Or. ADDRESS (Ciy, Stat., and ZIP COde)
PROGRAM PROJECT JTASK WORK UNIT

Washington, DC 20314-1000 ELEMENT NO. NO NO. CCESSION NO

11. TITLE (IncludeSecurCty Oasfication)

Rock Quality Designation (RQD) After Twenty Years

12. PERSONAL AUTHOR(S)
Deere, Don U., DonTIMEW.COVERED
Deere,1t3b. _ 114. DATE OF REPORT (Year, Mon-,h, Day) 115. PAGE COUNT
13a. TYPE OF REPORT
Final report FROMVI TO February 1989 100

16. SUPPLEMENTARY NOTATION

Available from National Technical Information Service, 5245 Port Royal Road, Springfield,
VA 22161.
17. COSATI CODES 18.SUBJECT TERMS (Continue on reverse if necessary and idertfy by block number)
FIELD GROUP SUB"ROUP Drill core Rock classification RQD
Fracture frequency Rock mechanics
In-S;itij modulIu• Rock alitY
by bock number)
19. ABSTRACT (Continue On reverse if necessary and identify
Twenty years of experience is now available in the use of the Rock Quality Designation (RQD)
in practice. The RQD is an index of rock quality in which a modified core recovery per
centage is obtained by counting only pieces of sound core 4-in. (100 mm) or greater in
length of NX size or larger core diameters. Experience now indicates both smaller (NQ) and
larger core diameters are appropriate; that slightly and moderately weathered core that can
not be hand broken be included; that length measurements be made along the center-line or
axis of the core piece; and that the requisite length of 4-in. (100 mm) be retained.
Problems with core breakage and loss occur in thinly bedded and schistose rocks, and,
particularly, with weak argillaceous rock interbedded with harder sandstone or limestone, a
problem that can be ameliorated by large diameter cores, shorter coring runs, and by use of
the best drilling equipment and techniques. Correlations of RQD with certain engineering
parameters are given, but the more recent classification system of Bieniawski or Barton
et al, which include the RQD as a parameter, are preferred for estimating the design -over)
20. DISTRIBUTION/AYAILABILITY OF ABSTRACT 21.ABSTRACT SECURITY CLASSIFICATION
0 SAME AS RPT. 03 DTIC USERS Unclassified
[I UNCLASSIFIEDUNLIMITED
22a. NAME OF RESPONSIBLE INDIVIDUAL
j
22b.TELEPHONE (include Area Code) 22€. OFFICE SYMBOL
are Obsolete.
PreviouS editions ~ ~ ECUI~IYLA IIAIN y 11 .'i
OF) IHMPAIQ"/.,;
DD Form 1473, JUN 96 Previouseditions are obsolete. Se URITY rIAC.IPF•ICAICN
UNCLASSI FIED

I
 UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE

6a. Name of Performing Organization

Don U. Deere, Consultant

6b. Address

Don U. Deere
6834 S.W. 35th Way
Gainesville, FL 32608

ROCKY MOUNTAIN CONSULTANTS, INC.

1960 Industrial Circle, Suite A
Longmont, CO 80501

19. ABSTRACT

-and construction parameters. For obtaining the RQD, the best drilling techniques
and prompt core logging in the field by a qualified engineering geologist or
geotechnical engineer should be used. The RQD is not a design parameter
that stands alone, but must be used together with an appreciation of the
detailed geology and the geotechnical aspects.

UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE
 PREFACE

This report was written by Don U. Deere, an independent

international consultant on dams, tunnels, and underground

powerplants, from Gainesville, Florida, and Don W. Deere,

geotechnical engineer and principal of Rocky Mountain

Consultants, Inc. of Longmont, Colorado.

The work was done under the sponsorship of the US Army Corps

of Engineers, Waterways Experiment Station (WES), Geotechnical

Laboratory (GL), Vicksburg, Mississippi, under Contract No.

DACW39 86M 4273, under date 86 SEP 02 and Amendment P00001 dated

86 OCT 22.

The project was conducted under the general supervision of

Dr. Don C. Banks, Chief, Engineering Geology and Rock Mechanics

Division; and Dr. W.F. Marcuson III, Chief, GL, WES. Col. Dwayne

G., Lee, EN, was the Commander and Director of WES. Dr. Robert

W. Whalin was the Technical Director.

u-" " r"'-:

S;...... ., ,~ - ~-,
~ J....----

C; 1 . 00/-
. . -. . - - -
 TABLE OF CONTENTS

PREFACE . . . . . . . . . . . . . . . . . . . . . i

PART I: INTRODUCTION. . . . . . . . .... • • • o . • 1

Background . . . . . . . .... 1
Purpose . . . . . . . . . . . . 1
Approach . . . . . . . . . . . .. . . 2

PART II: DEVELOPMENT OF THE RQD CONCEPT . . * • * . . . 3

The 1963 RQD Precursor . . . . . 3

The 1964-65 Developmental Period 6
The 1966-69 Testing Period . . .... 8

PART III: RECOMMENDED TECHNIQUES OF RQD LOGGING S. .. . 11

Core Diameter. . . .... 11

Length Measurement of Core Pieces. . 13
Assessment of Soundness. . . . . . .... 16
Length of Coring Run ..... ............ 19
PART IV: RQD CORRELATIONS. . . . . . .... 21
Tunnel Support/Reinforcement Design. 21
Prediction of In-Situ Modulus. .... . 26
Foundation Settlement Correlations . 30
Fracture Frequency . . . . . . . .... 31
Linear fracture frequency . . . . . . . . . . 32
Volumetric fracture frequency . . . . . . . . 32

PART V: UTILIZATION OF THE RQD IN LATER ROCK

CLASSIFICATION SYSTEMS. . . . . . . . . . . . . 35

Bieniawski's Rock Mass Rating System

(Geomechanics Classification). . . . . *. . . . . 35
Bartons Q System (Norwegian Geotechnical
Institute) . . . . . . . . . . . . . . . . . . . . 37

PART VI: PERTINENT QUESTIONS OF RQD IN PRACTICE . . . . 40

Mechanics of Determining RQD (1) . . . . . . . . 40

Core diameter (UA). . _ . : : 40

Length measurement of core'pieces (IlB) ... 42
,ength of coring run and of RQD
interval (1C) . . . . . . . . . . . . . 44

ii
 Special RQD Logging Problems (2) . . . .... . . 46

Drilling equipment and techniques (2A). 46

. 47
Prompt logging of core (2B) . ..
47
Applicability to various rock types (2C).
General problems . . . . . . . . . . . . 47
Shales, claystones, interbedded
sedimentary rocks. . . . . . . . . . . . 48
Limestone with solution cavities . . . . 49
Volcanics and metamorphics . . . . . . . 50

Orientation Effects (2D). . . . . . . . . . . 51

Desirability of Additional Geological

Observations (3) . . . . . . . . . . . . . . .* . . 52

Joint conditions (3A) . . . . . .... . . . 52

Local geology, weathering, fracture
. . . 52
frequency (3B). . . . . . . . . ....

Applications to Engineering and Construction (4) . 53

General (4A)-. . . . . . ... . . . . . . 53

Excavation, dredging, underwater
blasting (4B) . . . . . ... 54
Foundations, in-situ modulus (4C) 55
(4D). . . . . . . ....
55
Tunnels
Erosion resistance, roughness
coefficient (4E). . . ... . . . . . . 56

.. . ... 57
General Usefulness of RQD (5) . ...

Favorable experience (5A) . . .. . . . . . . 57

limitations (5B). . . . . . . . 59
Shortcomings,
. . . . . . 61
PART VII: CONCLUSION. . . . . . . . . ..

REFERENCES. . . . . . . . . . . . . . . . . . . . . . 64

APPENDIX A: COMMENTS AND QUESTIONS ON THE RQD FROM THE US

ARMY CORPS OF ENGINEERS

LIST OF TABLES

TABLE 1 - ROCK QUALITY CLASSIFICATION . . . . .. . . 22

TABLE 2 - GUIDELINES FOR SELECTION OF
PRIMARY SUPPORT FOR 20-FT TO
40-FT TUNNELS IN ROCK . . . . . . . . . . . 23
TABLE 3 - CORRELATIONS OF MODULUS RATIO WITH RQD AND
-VELOCITY INDEX. . . . . . . . .*. . . . . . 28
TABLE 4 - ALLOWABLE CONTACT PRESSURE q, ON
JOINTED ROCK. . . . . . . . . . *. . . .. 30

iii
LIST OF FIGURES

FIGURE 1 - RQD LOGGING. . . . . . . . . . . . . . . . . 12

FIGURE 2 - LENGTH MEASUREMENT OF CORE FOR RQD . . . . . 14
FIGURE 3 - ROCK QUALITY AND SUPPORT REQUIREMENTS
FOR TUNNELS OF VARYING DIMENSIONS. . . . . . 24
FIGURE 4 - RQD VERSUS MODULUS RATIO . . . . . . . . . . 27
FIGURE 5 - RQDVERSUS FRACTURE FREQUENCY . . . . . . . 33

iv
 PART I: INTRODUCTION

Background
The Rock Quality Designation (RQD) was developed as an index

of rock quality and was first used on a design and construction

job in 1964, and on two additional jobs in 1964-1965. Research

continued at the University of Illinois over the next several

years under the sponsorship of the US Air Force and US Department
of Transportation. The RQD was also used during this period on
several consulting jobs involving tunnels and shafts.
The publication that first brought the RQD to the attention

of the engineering and geology profession was the 1967 paper by

Deere and his colleagues at the University of Illinois (Deere et

al., 1967). The following year a chapter by Deere (1968) in the

rock mechanics book by Stagg and Zienkiesicz (1968) introduced

the concept to an international audience and led to its

acceptance and growing use in many countries.

The later rock classification systems for engineering

purposes developed by Bieniawski (1973) and by Barton et al.,

(1974) both include the RQD as an input parameter. Because these

systems are increasingly being used, there is interest in the RQD

logging techniques and applications.

Purpose

The purpose of the study was to evaluate twenty years of

experience by the senior author with the RQD in which many rock
types at a great number of engineering projects in many countries

have made their contribution. Conversations and correspondence

1
with many field engineering geologists and geotechnical engineers
have raised questions regarding the origin of the RQD, the

different techniques of measuring and logging, the optimal core

diameter and length of coring runs, and problems with core

breakage during drilling and handling, as well as questions

regarding its use in engineering design. The junior author

brings his own experience and questions after having logged a

variety of cores, managed geotechnical investigations, and

utilized RQD design correlations on many sites in the United

States and overseas over the past decade and a half.

Approach

The approach to the study was to organize the background

material relating to the early development in Part II, to discuss

and state the recommended RQD logging techniques in Part III, to

present and evaluate RQD correlations in Part IV, and to discuss

briefly in Part V the Bieniawski (1973) and Barton rock mass

classification systems (Barton et al., 1974).

Part VI is concerned with questions of RQD in practice as

posed by engineers and geologists of the US Army Corps of

Engineers and by some of their geotechnical consultants.

Conclusions are given in Part VII.

During the studies for this report, a shorter, preliminary

version entitled, "The Rock Quality Designation (RQD) Index in

Practice" was presented to the ASTM Symposium on Rock

Classification Systems for Engineering Purposes" (Deere and

Deere, 1988).

2
 PART II: DEVELOPMENT OF THE RQD CONCEPT

The 1963 ROD Precursor

During the development of rock mechanics in the United
States in the early 1960's, there was considerable interplay with
the European engineers and geologists and particularly with the
Salzburg School of Rock Mechanics in Austria, under the
leadership of Leopold Muller. Many of the senior author's

concepts were formulated from papers given and discussions held

at several of the annual Proceedings, Salzburg Colloquia on Rock
Mechanics. It was therefore with appreciation and satisfaction
that he accepted the invitation to author a paper for the first
volume of the new journal Felsmechanik und Ingenieuraeolocie
(Rock Mechanics and Engineering Geology) edited by Dr. Muller, an
outgrowth of the journal Geologie und Bauwesen.
That paper was entitled, "Technical Description of Rock
Cores for Engineering Purposes" (Deere, 1963). The ideas

presented were based not only on the background of the Salzburg

connection but also on the author's consulting practice in
foundation engineering, engineering geology and mining
engineering, and his program of graduate courses and research in
rock mechanics being developed at the University of Illinois.
The RQD concept was not presented in the 1963 paper as it
had not as yet been conceived in its entirety. However, the

important geological features that influence rock engineering

were emphasized, as was the information that could be obtained
from study of rock cores. A number of the passages are quoted in

the following paragraphs because (1) the published article is not

3
readily available, and (2) the ideas expressed formed a direct
precursor to the development of the RQD the following year:

"...Technical Description of Rock Cores. A careful

study of rock cores from boreholev can yield valuable
information concerning the nature of the in-situ rock
mass. The significant geological features are those
that influence the homogeneity of the rock mass and
include the occurrence of (1) surfaces of discontinuity
and (2) zones with materials of different hardnesses.
Detailed observations of these features should be made
and recorded on the boring logs. Complete and accurate
descriptions are necessary for rock mechanics studies
and for allowing the contractor to appraise the nature
of the in-situ rock and to plan and carry out his
construction procedure.
... Emphasis is given in this paper to those geological
features which can be observed in rock cores, and which
appear to the author to be significant in rock
engineering. The significant features include those
which have a direct bearing, almost overwhelmingly so,
on the homogeneity of the rock mass with respect to (1)
variations in hardness, and (2) physical discontinu
ities. The pertinent features when observed in the
rock cores should be carefully described and recorded
in the boring logs in such a manner so as to present a
factual record containing a minimum of interpretation.
From such boring logs interpretations may be made
concerning the character of the rock mass.
... Physical discontinuities are present in all rock
masses in the form of planes or surfaces of separation.
Geologically, these discontinuities are recognized as
joints, faults, bedding planes, or rock cleavage
planes. Terzaghi has referred to such features as
mechanical defects of rock. The permeability, shear
strength, and deformability of a rock mass are all
influenced by the number and kind of discontinuities
existing in the mass. Engineering projects involving
dam foundations, tunnels, underground chambers, and cut
slopes may be adversely affected unless the discontinu
ities are evaluated and their influence taken into
account during design and construction.
... A critical examination of rock cores can yield
valuable data concerning the occurrence and nature of
the mechanical defects in the rock mass from which the
cores were obtained. The various types of observations

that can be made are discussed in the following

paragraphs.

4
 The article goes on to describe types of discontinuities

(joints, bedding planes, cleavage planes, faults) and proposed

terminology for describing joint spacing and the thickness of

bedding units [e.g., moderately close joint spacing, 1 ft. - 3

ft. (30 cm - 1 m); and medium thick bed, also 1 ft. - 3 ft. (30

cm - 1 m)]. Probably of more significance in the development of

the RQD was the following statement:

... In describing the rock cores it is advised that the

length of the pieces of the core obtained in each
coring run be measured and recorded (e.g., 1 piece of
20 cm, 4 pieces of 10-15 cm, and 25 pieces of 2-10 cm,
etc.) These lengths are a direct response to the
spacing of the joints and fractures and the thickness
of the bedding. Unfortunately, they are also
influenced by the drilling method and technique.
Still, in the author's opinion, they are of sufficient
import to warrant describing.

Emphasis was then given to the importance of the surface

characteristic of the discontinuities and of filling materials:

... The behavior in an engineering project of a rock

mass transversed by discontinuities is probably more
influenced by the character of the joint surfaces and
the type of filling material along the discontinuities
than by the mere presence of the discontinuities.
Therefore in describing rock cores particular attention
should be given to those observations regarding the
tightness and irregularity of the surfaces as well as
to the kind of filling material between or along
adjacent surfaces.

It was noted that the degree of tightness could be described

as tight or Ope ; the degree of planeness by plane, curved, or

irregular; the degree of smoothness by slick, smooth, or rough;

5
and the infilling or altered materials as to thickness, t_•, and
hardness. The paper closes with discussion of lithology and
hardness, noting in particular the severe design and construction
problems that may arise from differences in hardness:

... Illustrative conditions are those encountered with

interbedded shales and limestones; with solution
widened and clay-filled joints, fault zones, and
bedding planes in limestone terrain; with altered and
weakened rocks along faults and shear zones in any type
of rock; and with the varied products of weathering in
the weathered rock zone where joint-block remnants
(often spheroidal) of fairly hard rock are surrounded
by soil-like material resulting from advanced
weathering and decomposition of the rock adjacent to
the joint. Many of these conditions will become
apparent during the geological mapping; however, the
extent of the condition can often only be determined by
means of boreholes. Consequently, the rock cores
should be studied with utmost consideration being given
to the detection of significant variations in hardness.

The 1964-65 Developmental Period

While consulting on the siting and design of a shaft,
tunnel, and chamber in granite in 1964 at the Nevada Test Site
for underground nuclear testing, it became clear to the senior
author that the site had poorer quality granite than an alternate
site. However, the detailed core logs prepared by well-qualified
geologists did not readily reveal the difference, perhaps because
of the emphasis on lithology, mineralogy, and alteration and the
lengthy descriptions of the jointing.

The ittributes of the core that visually indicated poor rock

conditions were the great number of small core pieces bounded by
weathered joints and sheared surfaces, the presence of numerous

6
rock fragments, and occasional core pieces of visibly altered

granite. By contrast, the rock cores from the alternate site

were of hard, nearly unweathered granite in core pieces of much
greater length. Even a casual examination showed that the

chemical alteration and the amount of jointing and shearing was

much less at the alternate site.
In an effort to illustrate this lesser alteration and

jointing at the alternate site as compared with the original site

and to be able to portray the rock quality graphically, it was

decided to use a "modified core recovery" procedure in which only

sound pieces of granite of 4 in. (100 mm) in length or longer

were counted. Thus, the quality of rock core was downgraded by

not counting the rock fragments, the pieces of core less than the

requisite length, pieces of altered granite, and unrecovered

core. The 4 in. (100 mm) requisite length was chosen after

considerable deliberation as being a reasonably lower limit for a

fair quality rock mass containing three or four joint sets of
close to moderate spacing.

The following day an oral and graphical presentation was

made to the designers and managers. Large-scale boring logs were

presented for each site with the "modified core recovery" plotted

with depth. Where this value was greater than 95 percent (later

changed to 90 percent) the interval was colored blue and was

designated as excellent quality rock; the 75-95 percent interval

colored green and designated aood quality; the 50-75 percent

interval colored orange and designated fair quality; the 25-50

percent and the 0-25 percent intervals colored red and

designated, respectively, Roor and very poor. The name Rock

7
Quality Designation (RQD) was applied to the overall procedure.

The visual display and associated descriptions were readily

assimilated by the audience and a rapid decision was made to
select the alternate site with the indicated good rock

conditions. Later construction of the facility corroborated the

generally good rock conditions at the selected site.

In the past, the percentage of core recovery had often been

used as an indicator of rock quality. However, with better

drilling techniques and with advancements in coring bits and

barrels, the percentage of core recovery was often nearly 100

percent even in closely jointed zones and fault zones.

Therefore, the "modified core recovery" and RQD concept offered

the possibility of a more valid technique of indexing the rock
quality for engineering purposes that took into account the

effects of fracturing, shearing, and alteration.

The concept was next tested on consulting projects in 1964

65 on highway tunnels on the Pigeon River for the North Carolina

Highway Department (quartzite, gneiss, and schist) and for the

foundation studies for the World Trade Center in New York on

massive schist and schistose gneiss. Meaningful results were

obtained on the delineation of rock zones of differing qualities

that resulted in substantial differences in design and

construction (personal files).

The 1966-69 TestinQ Period

The early success of the RQD on consulting projects

indicated that the concept was worthy of additional study and

research by the rock mechanics and engineering geology groups at
8
the University of Illinois. Sponsorship was obtained from the US
Air Force on the development of an engineering classification of
in-situ rock. Deere et al. (1969b) presented the complete report
of that investigation; some of the more pertinent results were
presented earlier (Deere et al., 1967; Deere, 1968; Hendron,

1968).

The 1967 reference by Deere and his colleagues at the

University of Illinois (Deere et al., 1967) was the first time

that the RQD concept had been presented in published form to the
engineering and geology profession' . The published work that
introduced the RQD to a wide international audience, and that no

doubt was responsible for its rapid acceptance in many countries,

was the 1968 book by Stagg and Zienkiewicz "Rock Mechanics in

Engineering Practice" that contained one chapter by Deere (1968)
and one by Hendron (1968) in which the RQD concept and
applications were discussed.
In the US Air Force studies, a number of sites were visited

and RQD measurements were made of existing rock core or of cores

from borings drilled specifically for that research project.
Different requisite lengths for core pieces to be counted for the

RQD were tried, as well as a weighting procedure. The weighting

procedure involved counting all pieces but giving

An incorrect reference was cited inadvertently in this paper

accrediting Deere with the introduction of the RQD in his
1964-paper "Technical Descriptions of Rock Cores for
Engineering Purposes." Actually, two mistakes were
involved. First, the date of the cited paper should have
been 1963, not 1964; and second, the RQD concept was not
given in that paper as it was not developed until 1964 and
was only available in file copies of consulting reports.

9
less weight to the smaller pieces by using the square of the

lengths for all pieces less than 1 ft. (300 mm). While this

eliminated the discontinuity at 4 in. (100 mm), it complicated

the procedure and the results did not appear appreciably better.

Therefore, the original 4 in. (100 mm) requisite length was

retained.

The Air Force study also included correlation with other

rock quality indices, with some in-situ rock properties, and with

tunnel supports and advance rates. A number of these

correlations are discussed in Part IV.

The US Department of Transportation in the late 1960's also

sponsored research at the University of Illinois on tunnel

support systems that included the RQD as a quality index for

predicting type and amount of required support (Deere et al.,
1969a; Peck et al., 1969; and Deere et al., 1970).

10
 PART III: RECOMMENDED TECHNIQUES OF RQD LOGGING

In this section several of the techniques for the RQD

logging of cores are reviewed. The procedures as given in the

original references (Deere et al., 1967; Deere, 1968) are

discussed together with some of the problems encountered and

modifications proposed by others or by the authors.

The RQD is a modified core recovery percentage in which all

the pieces of sound core over 4 in. long (100 mm) are summed and

divided by the length of the core run. The correct procedure for
measuring RQD is illustrated in Figure 1. The RQD is an index of

rock quality in that problematic rock that is highly weathered,

soft, fractured, sheared, and jointed is counted against the rock

mass. Thus, it is simply a measurement of the percentage of

"good" rock recovered from an interval of a borehole.

Core Diameter

The RQD was originally developed using NX-size core (2.155

in. or 54.7 mm diameter'). Deere (1968) specified that a minimum

NX-size core obtained with double-tube core barrels should be

used. This minimum size was specified to discourage a common

practice of the time of utilizing excessively small core sizes or

single barrel coring in geotechnical investigations; both of
which can result in poor recovery and excess core breakage.

Core diameters referred to in this report are nominal

dimensions. Actual diameters may vary slightly depending
upon core barrel manufacturer.

11
 LENGTH OF
L =10"
Z SOUND CORE > 4 INCHES (100mm.)
PIECES
ROD=
TOTAL CORE RUN LENGTH

10+ 7.5+8
RQD = x 1000%
LzO 48
HIGHLY WEATHERED wi
DOES NOT MEET m
SOUNDNESS REQUIREMENT C-)
RQD= 530/ (FAIR)

0
I-
w

CENTER LIN ~
PIECES<4
8HIGHLY WEATHERED Z
ROD
(ROCK QUALITY DESCRIPTION OF
DESIGNATION) ROCK QUALITY
0-25% VERY POOR
25-50 % POOR
50-75% FAIR
75-90 % GOOD
90-100% EXCELLENT
L 7.5"

L: 8"1

FIG. I- RQD LOGGING

 Experience in recent years indicates that diameters both
larger and slightly smaller than NX may be utilized for computing
RQD. The popular wire-line core, NQ (1.875 in., 47.6 mm

diameter) is now used extensively and in considered acceptable;

so are the larger HQ and PQ of the wire-line series and the 2-3/4
in., 4-in., and 6-in. sizes. The smaller BQ and BX sizes are

discouraged because of more potential core breakage and loss. If

they are used, a note should be made on the boring log indicating
that both core recovery and RQD may be slightly lower than if
taken on the preferred NQ size or larger. The topic of core

diameter is addressed in more detail in Part VI.

Variable length requirements for RQD measurement have been

proposed (Heuze, 1971). For example, instead of using the

standard 4-in. (100 mm) requisite length, a length equal to

double the core diameter was advocated (such as a 60-mm length

when using 30-mm diameter AX core). The authors believe that

4-in. (100 mm) requisite length should be used for all cases for

the purposes of standardization and comparison. Moreover, with

good drilling techniques, the lengths of the core pieces, at

center-line measurement, will be the same regardless of core

diameter since the spacing of natural unbonded joints does not

change.

Length Measurements of Core Pieces

The same piece of core could be measured three ways: along

the centerline, from tip to tip, or along the fully circular

barrel section (Fig. 2). No specific instructions were given in

13
 L

2" 0 CORE 2"0 CORE 20 0 CORE

4"0 CORE 4"0 CORE 4" 0 CORE

CENTERLINE TIP TO TIP FULLY CIRCULAR

A. CORRECT METHOD FOR CORE B. INCORRECT METHOD FOR CORE C. INCORRECT METHOD FOR CORE
LENGTH MEASUREMENT LENGTH MEASUREMENT LENGTH MEASUREMENT
- LENGTH OF-CORE -LENGTH DEPENDENT ON -LENGTH DEPENDENT ON
INDEPENDENT OF CORE DIA. CORE DIA CORE DIA.

FIG. 2 - LENGTH MEASUREMENT OF CORE FOR RQD.

Deere's previous papers (Deere et al., 1967; Deere, 1968; Deere*

et al., 1969b). The recommended procedure is to measure the core

length along the centerline (Fig. 1). This method is advocated

by the International Society for Rock Mechanics (ISRM),

Commission on Standardization of Laboratory and Field Tests

(1978, 1981). The center-line measurement is equivalent to a

scanline along the core axis. The reasons that the center-line

measurement is preferred are that 1) it results in a standardized

RQD that is not dependent on the core diameter, and 2) it avoids

unduly penalizing the quality of the rock mass for cases where
fractures parallel the borehole and are cut by a second set.
Core breaks caused by the drilling process should be fitted

together and counted as one piece. Drilling breaks are usually

evidence by rough fresh surfaces. For schistose and laminated

rocks, it is often difficult to discern the difference between

natural breaks and drilling breaks. When in doubt about a break,

it should be considered as natural, in order to be conservative

in the calculation of RQD for most uses. This practice would not

be conservative when the RQD is used as part of a ripping or

dredging estimate.

Some rocks, such as shales and claystones, often break up

into small discs or chips with time. Rock core with initial RQD

of 100 percent may break up in a period of hours or days into

core with zero RQD. This phenomenon is due to one or more

deleterioUs processes of slaking, desiccation, stress relief

cracking, or swelling. Thus, it is imperative that the RQD be

logged on site when the core is retrieved. The breakup of the

15
core over time, however, should be noted on the drilling logs, as
this is evidence of a rock property that may control design of a

structure.

Assessment of Soundness
Pieces of core which are not "hard and sound" (International

Society for Rock Mechanics, 1978, 1981) should not be counted for
the RQD even though they possess the requisite 4-in. (100 mm)
length. The purpose of the soundness requirement is to downgrade
the rock quality where the rock has been altered and weakened

either by agents of surface weathering or by hydrothermal

activity. Obviously, in many instances, a judgment decision must
be made as to whether or not the degree of chemical alteration is
sufficient to reject the core piece.

One procedure, which the authors have used, is not to count

a piece of core if there is any doubt about is meeting the
soundness requirement (because of discolored or bleached grains,
heavy staining, pitting, or weak grain boundaries). This
procedure may unduly penalize the rock quality, but it errs on
the side of conservatism. A second procedure which occasionally

has been used by the authors in recent years is to include the

altered rock within the RQD summed percentage but to indicate by

means of an asterisk (RQD*) that the soundness requirement has

not been met. The advantage of the method is that the RQD* will
provide some indication of the rock quality with respect to the

degree of fracturing, while also noting its lack of soundness.

Bieniawski (1974) addresses the soundness requirement as

follows:

16
 ... Since only hard, sound core is included in RQD
determination, this means that rock core which is
highly weathered receives zero RQD. For this purpose
"highly weathered rock" means that weathering extends
throughout the rock mass. The rock material is partly
friable, has no lustre and all material except quartz
is discolored or stained. Highly weathered rock can be
excavated with a geologist's pick...

The assessment of the soundness requirement merits further

consideration. There is no disagreement with Bieniawski's

suggestion that "highly weathered rock" receives zero RQD. Using

the weathering grades of the International Society for Rock

Mechanics (1978, 1981) (I-Fresh; II-Slightly Weathered; III

Moderately Weathered; IV-Highly Weathered; V-Completely

Weathered; and VI-Residual Soil), there is no doubt about Grade

I-Fresh being included and Grade VI-Residual Soil being excluded

from the RQD count. The remaining four categories all represent

degrees of weathering where judgment decisions must be made.

Grade II-Slightly Weathered is described as "Discoloration

indicates weathering of rock materials and discontinuity

surfaces. All the rock material may be discolored by weathering

and may be somewhat weaker externally than in its fresh

condition." Since the alteration is limited to discoloration,

possibly with somewhat lowering of strength, it appears logical

to accept this degree of "slightly weathered" Grade II in the RQD

count. The Grade V-Completely Weathered state by its very name

eliminates any core so described from the RQD count. Its

description is, "All rock material is decomposed and/or

disintegrated to soil. The original mass structure is still

17
largely intact" (International Society for Rock Mechanics, 1978,

1981).
The two remaining categories are III-Moderately and IV
Highly Weathered. The latter category, IV-Highly Weathered is

the one which Bieniawski (1974) eliminated from the RQD count.

The ISRM description is, "More than half of the rock material is

decomposed and/or disintegrated to a soil. Fresh or discolored

rock is present either as a discontinuous framework or as

corestones." Little (1969) in his description of residual

tropical soils uses the same terminology, Highly Weathered, and

states, "Rock so weakened by weathering that fairly large pieces

can be crumbled in the hands. Sometimes recovered as core by

careful rotary drilling. Stained by limonite." It is clear that

Highly Weathered rock should not be included in the RQD count

since it has been weathered to the point that it can be crumbled

in the hands.
The Grade III-Moderately Weathered category is described

(International Society for Rock Mechanics, 1978, 1981) as, "Less

than half of the rock material is decomposed and/or disintegrated

to a soil. Fresh or discolored rock is present either as a

continuous framework or as corestones." Little (1969) states for

Moderately Weathered rock, "Considerably weathered. Possessing

some strength; large pieces (e.g., NX drill cores) cannot be

broken by hand. Often limonite stained. Difficult to excavate

without the use of explosives." Because this category is close

to the borderline, it is of interest to consider another

description (Fookes and Horswill, 1970), "Term-Moderately

18
Weathered, Grade III, Abbreviation Mw...The rock is discolored;.

discontinuities may be open and surfaces will have greater

discoloration with the alteration penetrating inward; the intact
rock is noticeably weaker, as determined in the field, than the

fresh rock."

It is recommended that Grade III-Moderately Weathered rock

be accepted in the RQD count but that it be identified with an

asterisk as being less than sound. However, it possesses

sufficient strength, although moderately weathered, to resist

hand breakage of core pieces.

In summary, Grade I (Fresh) and II (Slightly Weathered) are

included in the RQD count, as is Grade III (Moderately Weathered)

but with the asterisk qualifier. Grades IV (Highly Weathered), V

(Completely Weathered), and VI (Residual Soil) are disregarded in

the RQD count.

Length of Coring Run

The RQD is sensitive to the length of the core run. For

example, a 11.8 in. (300 mm) long, highly fractured zone within a

massive rock would result in RQD values of 90 percent, 80

percent, and 40 percent, for respective run lengths of 12.9 ft.

(3 m), 4.9 ft. (1.5 m), and 1.6 ft. (0.5 m). Thus, the shorter

the run length, the greater the sensitivity of the RQD and, in

this case, the lower its value [becoming equal to zero for a 11.8

in. (300-mm) run encompassing the fractured zone).

The authors recommend that in general the calculation of the

RQD be based on the actual drilling-run length used in the field,

19
preferably, no greater than 5 ft. (1.5 m) and certainly not more

than twice that length. Actual length and nature of zones of

poor and good rock should be described in the drilling log and
could be supplemented by calculation of RQD on variable
"artificial run lengths" to highlight poor quality or good

quality zones. Many times this discrimination occurs naturally

in the drilling process; as zones of poor rock are encountered,

the run lengths are shortened to prevent blockage of the coring

bit and to enhance core recovery. The ISRM Commission on

Standardization of Laboratory and Field Tests (International

Society for Rock Mechanics, 1978, 1981) recommends RQD logging

using variable "run lengths" to separate individual beds,

structural domains, weakness zones, etc., so as to indicate any

inherent variability, and to provide a more accurate picture of

the location and width of zones with anomalously low or high RQD
values.

20
 PART IV: RQD CORRELATIONS

The original development of the RQD was for use in early

site evaluation to predict tunneling conditions. The difference

in rock quality could be easily visualized between borings and

between different depth zones. Zones of poor rock within a mass

could be easily "red flagged." Shortly after its development,

correlations were made between RQD and tunnel support

requirements. This application was expanded to correlate the RQD
with rock mass modulus and rock foundation settlement. Since

1970, the RQD has been used as a basic element of rock mass
classification systems. Subsequent correlations between RQD and

fracture frequency were made so that RQD or fracture frequency

could be theoretically calculated when only one of the parameters
was measured.

Tunnel Support/Reinforcement Design

The RQD was an early and simple method of classifying rock

masses for prediction of tunneling conditions and selection of
tunnel support. The US Army Corps of Engineers (1978) discuss

the use of the RQD method in their publication "Tunnels and

Shafts in Rock."

The RQD support criteria relate RQD and construction methods

to alternate support systems of steel sets, shotcrete, or

rockbolts. The method was developed utilizing numerous actual

consulting cases and published case histories. Detailed

discussions of the use of the RQD for tunnel support design are

21
provided in papers by Deere et al., (1969a, 1969b), Peck et al.,

(1969), Deere et al., (1970), Cecil (1970), Cording and Deere,

(1972), and Merritt (1972).

The RQD can be generally correlated to the common tunnelers'

classification as follows in Table 1 (after Deere et al., 1970):

TABLE 1
ROCK QUALITY CLASSIFICATION

Rock Quality ROD (t) General Tunnelers' Classification

Excellent 90-100 Intact

Good 75-90 Massive, moderately jointed

Fair 50-75 Blocky and Seamy

Poor 25-50 Shattered, very blocky and seamy

Very Poor 0-25 Crushed

Guidelines (Deere et al., 1970) for the selection of tunnel

support/reinforcement systems based on the RQD for tunnels

between 20 and 40 ft. (6.1 m and 12.2 m) in diameter are given in

Table 2. Reduced support is shown for machine bored tunnels over

conventionally excavated drill-and-blast tunnels due to less rock

disturbance.

Merritt (1972) prepared correlations between RQD and

required tunnel support (no support or local bolts; pattern

bolting; or steel sets) for various sized tunnels. His

correlation is presented in Figure 3, together with the

recommendations of Cecil (1970) which include shotcrete.

Although the RQD does not take directly into account

important rock mass characteristics such as joint infillings,

22
 TABLE 2

GUIDELINES FOR SELECTION OF PRIMARY SUPPORT FOR 20-FT TO 40-FT TUNNELS IN ROCK

Alternative Support Systems

Rock Bolts(a)
Construction Steel Sets (Conditional use in poor and very Shotcrete (b)
Rock Quality Method poor rock) (Conditional use in poor and very poor rock)

Rock Load Spacing of Additional Total Thickness

(B-Tunnel Weight Pattern Requirements and -------------------------------- Additional
Width) ofSets Spacing(c) Bolts Anchorage Limitations(a) Crown Sides Support (b)
S................................................................................................................................................................................
Excellent(d) Boring (0.0-0.2)B Light None to None to Rare None to occasional None None
RQD • 90 machine occasional occasional local application
Drilling and (0.0-0.3)B Light None to None to Rare None to occasional None None
blasting occasional occasional local application
2 to 3 in.
Good(d) Boring (0.0-0.4)B Light Occasional Occasional Occasional mesh and Local application None None
RQD - 75 machine to 5 to 6 ft to 5 to 6 ft straps 2 to 3 in.
to 90 Drilling and (0.3-0.6)6 Light 5 to 6 ft 5 to 6 ft Occasional mesh and Local application None None
blasting straps 2 to 3 in.
Fair Boring (0.4-1.O)B Light to 5 to 6 ft 4 to 6 ft Mesh and straps as 2 to 4 in. None Provide for
ROD - 50 machine sedium required rock bolts
to 75 Drilling and (0.6-1.3)6 Light to 4 to 5 ft 3 to 5 ft Mesh and straps as 4 in. or more 4 in. or more Provide for
blasting medium required rock bolts
Poor Boring (1.0-1.6)8 Medium 3 to 4 ft 3 to 5 ft Anchorage may be hard 4 to 6 in. 4 to 6 in. Rock bolts am
RQD - 25 machine circular to obtain. Considerable required (approx.
to 50 mesh and straps required. 4-6 ft cc.)
Drilling and (1.3-2.0)B Medium 2 to 4 ft 2 to 4 ft Anchorage may be hard 6 in. or more 6 in. or more Rock bolts as
blasting to heavy to obtain. Considerable required (approx.
circular mesh and straps required. 4-6 ft. cc.)
Very Poor Boring (1.6-2,2)B Medium 2 ft 2 to 4 ft Anchorage may be impossible. 6 in. or more on whol a section Medium sets as
ROD < 25 machine to heavy 100% mesh and straps re required
(excluding circular quired.
squeezing
and swelling Drilling and (2.0-2.8)B Heavy 2 ft 3 ft Anchorage may be impossible. 6 in. or more on whole section Medium to heavy
ground) blasting circular 100% mesh and straps re sets as required
quired.
Very Poor, Both up to 250 ft Very 2 ft 2 to 3 ft Anchorage may be impossible. 6 in. or more on whole section Heavy sets as
squeezing or methods heavy 100% mesh and straps as re required
swelling circular required.
ground

NOTE: Table reflects 1969 technology in the United States. Groundwater conditions and the details of jointing and weathering
should be considered in conjunction with these guidelines particularly in the poor quality rock. See Deere et. al. (1969a)
for discussion of use and limitations of the guidelines for specific situations.

a Bolt diameter - 1 in. length - 1/3 to 1/4 tunnel width. It may be difficult or impossible to obtain anchorage with mechanically
anchored rock bolts in poor and very poor rock. Grouted anchors may also be unsatisfactory in very wet tunnels.
b Because shotcrete experience is limited, only general guidelines are given for support in the poorer quality rock.
c Logging requirements for steel sets will usually be minimal in excellent rock and will range from up to 25 percent in good rock
to 100 percent in very poor rock. /
d In good and excellent quality rock, the support requirement with in general be minimal but will be dependent on joint geometry,
tunnel diameter, and relative orientations of joints and tunnel.
roughness, orientation, or state of stress, the authors believe

it is still a useful tool in predicting ground conditions and

support requirements for tunnels. The analysis is quick and

inexpensive and may be used as the sole design method for

experienced tunnel engineers or geologists or simply be used as a
check guide on other more comprehensive design methods.

The authors have encountered cases where the RQD did not
correlate well with required tunnel support. For example, there

are cases where the RQD was in the good to excellent range yet
considerable tunnel support was required. Two examples of this

are as follows:

o open solution-widened joints in a massive limestone.

Tunnelling through this rock mass resulted in shifting

of large rock blocks bounded by the solutioned joints.

o Small clay-filled joints or shears that were moderately

spaced but adversely oriented within a generally
massive rock mass. Tunneling through this rock

resulted in very blocky ground requiring more support

than predicted by RQD.

Cases where very low RQD's have overpredicted support/

reinforcement requirements include the following:

o Highly fractured volcanic (or granitic) rocks -- the

joints or fractures were rough, tight, discontinuous,

.well-interlocked, and under sufficient stress to

prevent loosening.

25
 Prediction of In-Situ Modulus

Another aspect of the University of Illinois RQD research in

the late 1960's was the correlation of the RQD (or velocity
ratio) with the in-situ modulus of deformation. Obviously, the

greater the fracturing and alteration the lower the RQD and,

also, the lower the modulus; correlations showed this to be true

(Deere et al., 1967; Hendron, 1968; Deere et al., 1969b; Coon and

Merritt, 1970).

The modulus or elasticity of an intact rock specimen can be

measured in the laboratory by plotting stress versus strain in

the unconfined compression test. The laboratory modulus of an

intact rock specimen may be defined by Et 50 (the slope of a line

tangent to the stress-strain curve taken at 50 percent of failure

stress).

This modulus of elasticity in the lab is, of course, higher

than the modulus of deformability (static modulus of elasticity)

of the rock mass because the in-situ rock mass has discontinu

ities or joints. The ratio of intact lab modulus (Ets 0 or Elab)

to the in-situ rock mass modulus (Ed or Erk mm) correlates in

a general way with the RQD.

The relationship of the modulus ratio to the RQD as

developed by Coon and Merritt (1970) is shown on Figure 4. The

correlation is useful above RQD>60 percent but insufficient data

exist for the low RQD range.

The junior author has used this correlation on small

projects where it is not economically feasible to perform large

scale in-situ testing. For example, the correlation has proved

26
 U)
U)
a
E

w0

0
I

.-J

ROCK QUALITY DESIGNATION , ROD , %

0 DWORSHAK DAM,GRANITE GNEISS, SURFACE GAGES

BURIED
0
0 TWO FORKS DAMSITE , GNEISS
YELLOWTAIL DAM , LIMESTONE
GLEN CANYON , SANDSTONE
AFTER COON a MERRITT
1970

FIG. 4 - RQD. VERSUS MODULUS RATIO

quite useful in evaluating the safety of existing concrete dams
subject to loading from flood-overtopping, where the modulus of
deformation of the foundation is required as an input factor.
Drill cores of the rock foundation are tested for lab moduli and

then corrected by the modulus ratio reduction factor.

The Velocity Index can be substituted for RQD where RQD

information is not available or only limited borings are

performed. The Velocity Index is defined as the square of the
ratio of field in-situ seismic compressional velocity (Vf) to the
laboratory compressional sonic velocity (V1 ) or (Vf/Vl) 2 . The
field velocities may be taken via seismic refraction, cross-hole,
or downhole techniques. Laboratory sonic velocities are measured
on core specimens loaded to 3000-psi (211-kg/cm2 ) stress levels.
The correlation is nearly one to one between RQD and Velocity
Index. Table 3 displays the correlation between RQD, Velocity
Index, and the Modulus Ratio (after Coon and Merritt, 1970).
TABLE 3
CORRELATIONS OF MODULUS RATIO WITH RQD
AND VELOCITY INDEX
Velocity Modulus Ratio
Classification Ro Index Ek-mas-•.
Very Poor 0 - 25 0-0.20 < 0.20
Poor 25 - 50 0.20-0.40 < 0.20
Fair 50 - 75 0.40-0.60 0.20-0.50
Good 75 - 90 0.60-0.80 0.50-0.80
Excellent 90 - 100 0.80-1.00 0.80-1.00

Kulhawy (1978) expounded on the RQD/Modulus Ratio or

Reduction Factor by including an additional parameter of joint

stiffness. He developed a series of curves that allowed

estimation of the modulus reduction if RQD was measured and joint

28
stiffness was estimated from a table of representative measured

values for different rock types. The estimated modulus was then

proposed for use in a rational estimation of settlement on rock

foundations.

Bieniawski (1978) proposed a method for estimating the

modulus of deformation based on his RMR ratings of which RQD is a

parameter. His preliminary work showed reasonably good

correlation. The correlation was given as:

Emck ma ' 2 X RMR - 100 (1)

where Emck ma " in-situ static modulus of

deformation in GPa

RMR = rock mass rating in accordance with the

Geomechanics Classification - for RMR >50.
and Erok mas = 1 0 (RMR-10)/40 for RMR <50 (2)
(Serafim et al., 1983)

The senior author over the last decade has not used the RQD

correlation extensively but has employed for preliminary

estimates the unpublished correlation of seismic P-wave velocity
or seismic modulus and the in-situ modulus (Ei,•JE,t,tiC of

rock ma" = 1.5 to 10, often 4 to 5), or the correlation with the

shear wave frequency of Schneider as given by Bieniawski (1978).

For critical cases, the authors prefer large-scale in-situ

testing where the loading direction in the testing approximates

that in the prototype structure so that the significant rock

joints can be appropriately tested.

In summary, the RQD - Modulus Reduction correlation is

useful for obtaining a rough estimate of in-situ modulus of

deformation. The authors recommend that it be used in

conjunction with other modulus estimating techniques.

29
 Foundation Settlement Correlations
Peck et al., (1974) proposed allowable foundation contact

pressures on jointed rock on the basis of the RQD. They stated

with respect to settlement on unweathered rock:

Unless the strength of the intact rock is extremely

low, roughly equal to or less than that of plain
concrete, the allowable contact pressure beneath
foundations is governed exclusively by the settlement
associated with the defects in the rock, and not by
strength. The compressibility is closely related to
the spacing and direction of the joints, whether they
are tight or open, and whether they are filled or
coated with softer materials. If the joints are tight
or are not wider than a fraction of an inch, the
compressibility is reflected by the RQD (Art. 5.3).

Table 4 presents the maximum allowable contact pressure for

different RQD values for a total maximum settlement of 0.5 in.
(12.7 mm) proposed by Peck et al., (1974). The tabulated value

of allowable pressure should not be used if it exceeds the

unconfined compressive rock strength.

TABLE 4
ALLOWABLE CONTACT PRESSURE q. ON JOINTED ROCK

(tons/sQ ft) (lb/sa in)

100 300 4170
90 200 2780
75 120 1660
50 65 970
25 30 410
0 10 140

Peck (1976) in a later paper presents the same table and

notes, "...This correlation can give useful results if tempered

30
by a detailed knowledge of the geology of the deposit. However,

it is only a first crude step..."

Kulhawy (1978) reviewed this concept in his paper and agreed

that "It does provide a convenient starting point for evaluating

foundations on rock masses." He proposed a method for providing

a quantative estimate of rock foundation settlement based on RQD,

Modulus Reduction or Ratio, and joint stiffness, as previously

noted.

Fracture Freauency

There are instances where it is useful to convert fracture

frequency to RQD or vice versa. For example, use of rock mass

classification systems usually require RQD as an input, and if

RQD has not been directly measured in core, measurement of

fracture frequency from outcrop scanline surveys can be

converted. Various conversion factors have been derived and are

discussed below.

A word of caution should be noted; there is not a direct or

totally appropriate conversion available since RQD is a much more

general measure of rock mass or rock core quality than fracture

frequency. RQD discounts for core loss and highly weathered or

soft rock zones (soundness requirement). Fracture frequency and

RQD are closely related, however, for an unweathered rock mass

that is degraded only by fracturing.

Therb have been various publications that discuss the

fracture frequency/RQD correlation. Chronologically these

include, Deere et al., (1967); Priest and Hudson (1976);

31
Kulhawy (1978); Goodman and Smith (1980); Wallis and King (1980);
Priest and Hudson (1981); Palmstrom (1982); Hudson and Priest
(1983); Sen (1984); Sen and Kazi (1984); Kazi and Sen (1985).

Linear fracture freguency

Priest and Hudson (1976) define the correlation between

linear fracture frequency and RQD as randomly distributed

RQD = 100e -O.1? (0.1N+ 1) (3)

where A = mean discontinuity frequency per meter.

or

RQD = 110.4 - 3.68 ?, when 6 < X < 16 (4)

These appear to be reasonable correlations for use in rock

outcrop scanline surveys and were confirmed by 27 surveys in

chalks and mudstones. This randomly distributed theoretical

relationship was also confirmed in the field by Wallis and King

(1980) at a site within granite.
Goodman and Smith (1980) explored the theoretical models and

bounds of possibility in the correlation. The relationships

between RQD and fracture frequency are shown in Figure 5. The

Priest and Hudson relationship (1976), as presented above,

approximates the line of averages shown in Figure 5 and appears

to be a preferred formula.

Volumetric fracture freguency

Palmdtrom (1982) developed the concept of volumetric joint

count (Jv). This is a simple measure of the degree of jointing

and refers to the number of joints per cubic meter. Palmstrom

32
 100

75

6
50
0a,

25

0 15 20 25 30 35

DISCONTINUITIES PER I.5m. RUN

AFTER GOODMAN 8 SMITH

1980

FIG. 5 - RQD. VERSUS FRACTURE FREQUENCY

33
derived the theoretical correlation between Jv and RQD as
RQD = 115 - 3.3 (Jv) for Jv > 4.5 (5)

This conversion factor has been advocated by Barton et al.,

(1974) for use in the Q System Rock Mass Classification as well
as by the International Society of Rock Mechanics (1978).
Kazi and Sen (1985) proposed a rock mass parameter termed

volumetric RQD (V. RQD). The V. RQD is calculated by summation

of the volumes of intact blocks bigger than 0.001 m3 [i.e., for a
cubic block with 4 in. (100 mm) sides] divided by the total rock
mass volume, expressed as a percentage.
The volumetric joint count (Jv) and volumetric RQD (V. RQD)
are refinements on linear fracture frequency and RQD that account

for 3-dimensionality of discontinuities. They may prove useful

in cases where oriented coring and 3-D rock exposures are

available for inspection and when one has case history experience
with these parameters.

34
 PART V: UTILIZATION OF THE RQD IN LATER ROCK
CLASSIFICATION SYSTEMS

system
In the early 1970's a number of rock classification

were introduced. Two that have gained international acceptance

Bieniawski
and that are increasingly being used are those of

(1973) and Barton et al., (1974). Both systems use the RQD as an

input parameter.

Bieniawskil's Rock Mass Rating System (Geonechanics

Classification)

Bieniawski (1973) in the presentation of his classification

of jointed rock masses notes:

7 has a
... Deere's very practical and simple approach
considerable potential in relating 21his rock alitYin
desiqnation (RQD) to tunnel support as well as
22. However,
estimating deformability of rock masses
the RQD method disregards the influence of joint
may
orientation, continuity and gouge material which
all be of great importance in some cases.

Bieniawski proposed (1973) that for his Geomechanics

be incorporated:
Classification the following parameters should

o Rock Quality Designation (RQD);

o State of weathering;
o Uniaxial compressive strength of intact rock;
o Spacing of joints or bedding;
o strike and dip orientations;
o separation of joints;
o Continuity of joints; and
o Groundwater inflow.

With respect to the RQD he notes:

35
 ... The state of the rock cores recovered in a drilling
program is a valuable indication of the in-situ
condition and probable engineering behavior of a rock
mass1 °. Various criteria may be used for quantitative
description of the rock quality in the cores, such as
core recovery, fragment size, fracture frequency or
rock quality designation (RQD). While the actual
choice is largely a matter of personal preference, the
Author advocates the use of RQD because it has been
found particularly useful in classifying rock masses
for selection of tunnel support systems 2 , 20, 21.
... It should be noted that for RQD determination, core
of at least 50 mm in diameter should be used and double
tube N size core barrels (75 mm OD) with non-rotating
inner barrels are strongly recommended 24 .

With respect to the limitations of RQD and its practical

advantages Bieniawski (1973) states:

... The limitations of the RQD are that is disregards

the influence of joint orientations, continuity and
gouge material. On the other hand, the RQD procedure
is simple, inexpensive and reproducible. As a result
it is used extensively in the U.S.A. and Europe and is
currently quickly gained acceptance in South Africa.
... If the RQD method is used in core logging, there is
no need for determination of also fracture frequency
(i.e. the number of fractures over an arbitrary length)
or the fragment size, as this would be duplication of
effort.

The RQD parameter was given a rating in the classification

procedure which in the 1973 article ranged from 3 for very poor
rock to 16- for very good rock. Ratings for the other parameters
were obtained and summed to give the total rating. The value of
the total rating then defined the rock mass class; for example, a
rating of 70 to 90 indicated Class No. 2, good rock.

36
 Over the years, as more experience was obtained, several

changes were made to the classification system. The reader is

referred to Bieniawski's recent paper given at the ASTM Symposium

in
on Rock Classification Systems for tngineering Purposes held
June 1987 in Cincinnati (Bieniawski, 1988) for current

classification procedures and utilization of the method. The

term Rock Mass RatinQ System (RMR) appears to be gaining

preference over the previous name of Geomechanics Classification.

Barton's 0 System (Norwegian Geotechnical Institute)

A group from the Norwegian Geotechnical Institute (NGI)

proposed an engineering classification of rock masses for the

design of tunnel support (Barton et al., 1974). The rock mass

quality Q was proposed, being the result of six classification

parameters: the RQD index, the number of joint sets Jn, the
or
roughness of the weakest joints Jr, the degree of alteration
filling along the weakest joints Ja, the degree of water inflow

Jw, and a stress reduction factor SRF. With respect to the RQD

they note:

... The RQD index happens to be one of the better single

parameters since it is a combined measure of joint
frequency and degree of alteration and discontinuity
fillings, if these exist. However, it is relatively
insensitive to several important properties of rock
masses, in particular the friction angle of altered
joint fillings (Cording and Deere, 1972), and the
roughness or planarity of joint walls.

Barton el al., (1974) modified the RQD by multiplying the

indicators of:
value by appropriate factors that were considered

37
 1. Relative block size, (RQD/Jn)
2. Inter-Block shear strength, (Jr/Jn)
3. Active stress, (Jw/SRF)
The overall quality Q is given by the product:
Q = (RQD/Jn) (Jr/Ja) (Jw/SRF) (6)
Nine ranges of Q were identified with the following
descriptive terminology (eliminating herein the lowest and
highest ranges for simplicity and as being of lesser interest):

Extremely poor, 0.01-0.1; Very poor, 0.1-1; Poor, 1-4; Fair, 4

10; Good, 10-40; Very good, 40-100; and Extremely good, 100-400.

Barton et al., (1974) stated that when borecore was

unavailable, the RQD could be estimated as follows (Barton quotes

personal communication with Palmstrom, 1974):

RQD- 115 - 3.3 Jv (7)
where Jv - total number of joint per m3

(RQD - 100 for Jv < 4.5)

For details of the methodology and the application to

selecting tunneling support the reader is directed to the recent

publication by Barton (1988).

Hoek and Brown (1980) in their book state:

The large amount of information contained in . . . (the

instruction table] . . . may lead the reader to suspect
that the NGI Tunneling Quality Index is unnecessarily
complex and that it would be difficult to use in the
analysis of practical problems. This is far from the
case and an attempt to determine the value of Q for a
typidal rock mass will soon convince the reluctant user
that the instructions are simple and unambiguous and
that, with familiarity . . . [the tables] . . . become
very easy to use.

38
 Bieniawski (1976) made a comparison of his RMR ratings and

Barton's Q rock quality for 111 cases and found a reasonably good
correlation with the following relation:
RMR = 9 ln Q + 44 (8)

Bieniawski (1976) states, ". . . the author has found the

NGI system is relatively easy to apply once the user is fully

familiarized with its principles." He recommends that both the

Geomechanics Classification (RMR) and the NGI Classification (Q)

be used on each project for cross-checking purposes. The authors

agree that this is worthwhile in order to accumulate experience,

not only in the correlation of the two classifications but also

in the correlation with design parameters and construction
experience.

39
 PART VI: PERTINENT QUESTIONS OF RQD IN PRACTICE

Each District and Division of the US Army Corps of Engineers

submitted comments and questions about the RQD to Dr. Don C.
Banks, Chief, Engineering Geology and Rock Mechanics Division,
Geotechnical Laboratory, US Army Waterway Experiment Station,
Vicksburg, Mississippi. A total of 28 letters were received in
December 1986 and January 1987 from the Corps as well as from a
number of geotechnical consultants to the Corps.
The authors of this report have placed each question or
comment into one of five general categories that have been
established as follows:
1. Mechanics of Determining RQD,
2. Special RQD logging Problems,
3. Desirability of Additional Geological Observations,

4. Applications to Engineering and Construction, and

5. General Usefulness of RQD.

The categorized questions and comments as excerpted from the

28 letters of response are presented in the Appendix. This

section presents the authors' replies. The number following each

topic heading corresponds to the outline in the Appendix.

Mechanics of Determining ROD (1)

Core diameter (1A)

The topic of core diameter proved to be one of the more

popular subjects for comments and questions. Thirteen responses

addressed this item. The principal question was if diameter

40
larger or smaller than NX size could be used and if correlation

coefficients would be necessary. Several Corps of Engineer

Districts use NQ, HQ, and PQ wire-line coring and/or 4 in. (100

mm) coring.
As noted in Part III, the original work on RQD was done

almost exclusively on NX-size core. Deere (1968) recommended

that cores of at least NX size obtained by double-tube core

barrels be used together with proper drilling supervision.
Experience of the last decade has shown that the wire-line

series of core bits and barrels is increasingly being used,

particularly for the deeper holes. The NQ core of 1-7/8-in.

diameter' (1.875 in., 47.6 mm) is now perhaps as common as NWX

(or NWM) size (2.155 in., 54.7 mm) and the RQD is being taken on

either size, interchangeably without any correlation coefficient,

which appears to be acceptable practice.

The question then arises as to the next size smaller in both

categories [BWX, or BWM, of diameter 1.655 in. (42.0 mm); and BQ,

of diameter 1-7/16 in. (1.438 in., 36.5 mm)]. Experience has

shown that in good quality rock these sizes give similar results
to those obtained with the larger sizes. However, in weathered

and heavily fractured rock, and in weak sedimentary and foliated

and schistose metamorphic rocks, there is a tendency for more

core breakage and, perhaps, more core loss. Attempts can be made

to fit the core breaks back together for core measurements to

Core diameters referred to in this report are nominal

dimensions. Actual diameters may vary slightly depending
upon core barrel manufacturers.

41
compensate for some of the breakage. The authors believe that

the RQD should be taken on the BWX and BQ cores but that a note
should be added to the boring log pointing out that both core

recovery and RQD values may be slightly lower than if taken with

the recommended NQ size or larger. The AWX-AWM and AQ sizes

(1.185 in., 30-mm diameter; and 1-1/16 in., 27.0 mm,

respectively) are considered too small to be used for RQD because

of their potential for core breakage, grinding, and loss.

At the other end of the scale the larger diameter HQ (2-1/2

in., 63.5 mm); the 2-3/4 in. (2.690 in., 68.3 mm); the PQ (3

11/32 in., 3.343 in., 85.0 mm); the 4 in. (3.970 in., 100.8 mm);

and the 6 in. (5.970 in., 151.6 mm) are all acceptable for the

RQD. The HQ and 2-3/4-in. sizes are quite common now,

particularly for the upper portion of a borehole. In using the

larger diameter cores the RQD measurements must be taken along

the core axis centerline as described in the following section.

The 4-in. (100 mm) requisite length for a core piece to be

counted still would apply.

Length measurement of core pieces (1B)

Included among the ten comments and questions received on

this topic were three different items: (1) the position of the

measurement, (2) the recommendation (by others) of using a

requisite length of twice the core diameter, and (3) the problem

of distinguishing between natural and induced fractures. A

discussion of these items is also presented in Part III.

42
 The original RQD papers by Deere and his colleagues at the

University of Illinois never specifically outlined where the

length measurement should be taken on a core and thus
clarification is needed. Experience has shown that the length

should be measured at the core axis or centerline advocated by

the International Society of Rock Mechanics (1978, 1980). This

method of measurement is equivalent to a scanline and thus is

independent of core size and is less sensitive to joint angle.

Confusion as to the requisite length measurement arose due

to publications by others (Heuze, 1971; Goodman, 1981) that

defined RQD as "percentage recovery of core in lengths greater

than twice its diameter." This statement is approximately

correct for the N-sized core only. The 4-in. (100 mm) requisite

length, measured at centerline, should be used for all applicable

core sizes. The spacing of natural unbounded joints does not

change with core size.

It is often very difficult to distinguish between natural

and induced fractures. A committee of the International Society

of Rock Mechanics (1978, 1981) addressed the problem as follows:

.. When estimating frequency or RQD from drillcore it

is necessary to discount fresh artificial breaks (frac
tures) clearly caused by the drilling process, and also
those made deliberately when fitting core into the core
boxes. The following criteria are suggested:

(i) A rough brittle surface with fresh cleavage

planes in individual rock minerals indicates
an artificial fracture.

(ii) A generally smooth or somewhat weathered

surface with soft coating or infilling
materials such as talc, gypsum, chlorite,

43
 mica or calcite obviously indicates a natural
discontinuity.
(iii)In rocks showing foliation, cleavage or
bedding it may be difficult to distinguish
between natural discontinuities and
artificial fractures when these are parallel
with the incipient weakness planes. If
drilling has been carried out carefully then
the questionable breaks should be counted as
natural fractures, to be on the conservative
side.
(iv) Depending upon the drilling equipment part of
the length of core being drilled may
occasionally rotate with the inner barrels in
such a way that grinding of the surfaces of
discontinuities and fractures occurs. In
weak rock types it may be very difficult to
decide if the resulting rounded surfaces
represent natural or artificial features.
When in doubt the conservative assumption
should be made, i.e. assume that they are
natural.
(v) It may be useful to keep a separate record of
the frequency of artificial fractures (and
associated lower RQD) for assessing the
possible influence of blasting on the weaker
sedimentary and foliated or schistose
metamorphic rocks.

Length of corinQ run and of ROD interval (1C)

A total of five questions or comments were received

regarding the appropriate interval or run length over which to
measure the RQD. This concept is discussed in Part III with
further discussion below.
The RQD is highly sensitive to core-run length or interval,
providing more delineation of anomalously "poor" or "good" rock
zones with-shorter lengths. The authors advocate the-following

procedure:

44
1. Log RQD as the core comes out of the ground based on
the actual drill-run lengths and record on the drilling
logs. The length of coring-run should preferably not
exceed 5 ft. (1.5 m) but in more massive rocks where

recovery is 100 percent, 10-ft. (3 m) runs are

acceptable.

2. During drilling, the actual length of poor and good

rock zones should be described by prose in the drilling
log and should be supplemented by calculation of RQD on

"variable artificial run lengths" to highlight poor

quality or good quality zones, changes in lithology,

etc. This is analogous to the standard practice of

performing packer permeability tests on 20-ft. (6 m)

intervals within a boring, followed up by select tests

on smaller intervals in areas of high water "takes."

3. After one has gathered the proper information in the

field by the logging procedures described above, the

RQD values can then be assembled for different areas,

depths, etc. by calculation of weighted averages. For

example, a weighted average for RQD can be calculated

for each boring to compare one area of the site with

another. The weighted average of RQD for core taken

-within 2 or 3 tunnel diameters of a tunnel alignment

could be assembled and used for classification of the

tunnelling rock. Weighted averages can also be

calculated for each rock type encountered, each

structural domain, and for the upper weathered zone.

45
 Special ROD LoQginQ Problems (2)

Drilling eciuipment and technigues (2A)

Three comments were submitted with respect to the
questionable reproducibility of RQD since it is dependent on

human factors (skill and attitude of the drill operator) and on

the equipment used. The authors agree that these can be a

serious problem. It is important in all aspects of geotechnical

investigations to obtain the best information possible, not just

for RQD measurement. The engineering geologist or geotechnical

engineer can reduce the influence of these operational factors

by: (1) specification of proper drilling equipment in the bid
documents (e.g., double- or triple-tube core barrels, etc.) and

(2) providing for trained technical supervisory personnel on-site

during drilling.

In addition, correct measurement of RQD calls for dis

counting mechanically induced core breaks, although there are not

always easily discerned. The junior author has had some success

reducing drilling breakage and core loss by specifying a two

tiered payment system, whereby footage was paid for at one scale

for recovery above 95 percent, and at a second lower scale for

poorer recovery. This system of payment helps counter the

traditional system whereby the drill operator receives daily

footage bonuses from his company, which may result in overzealous

hole advance.

46
Prompt logging of core (2B)

The problem of core deterioration with time and handling was

pointed out, especially when drilling thinly bedded argillaceous

rocks.

The authors advocate logging of the core immediately after

it is removed from the ground. Valuable information is lost

every time the core is handled. This requirement is paramount

when dealing with shales that undergo time-dependent slaking,

desiccation, stress-relief cracking, or swelling.

Applicability to certain rock types (2C)

By far, the greatest number of comments and questions

received were with respect to the applicability of RQD to certain

rock types. A total of 16 comments were received on this subject

and are tabulated in the Appendix, Section 2C. The comments have

been subdivided into (1) General Problems; (2) Shale, Claystone,

Interbedded Sedimentary Rocks; (3) Limestone with Solution

Cavities; and (4) Volcanics and Metamorphics.

General problems. Six comments or queries were received

that have been lumped together under this heading. The princi

pal question appears to be if the RQD procedure is applicable to

all rock types. Yes, it has been applied to all lithologies.

Difficulties often arise with thinly bedded, laminated

sedimentary rocks and schistose or foliated metamorphic rocks.

Such rocks are prone to breakage along the incipient-weak

surfaces during drilling and handling. Good drilling techniques

with minimum vibration and large diameter cores (HQ or larger)

47
can yield intact cores. These must be logged immediately before
they break up due to handling, drying, and stress-relief

cracking. When drilling subparallel to the weakness direction,

it is very difficult to obtain cores without breakage or core

loss. Boreholes should be drilled at various orientations to

investigate the directional sensitivity.

Another problem is the artificial discontinuity in the RQD
count at the 4-in. (100 mm) requisite length. Such a break
penalizes too heavily the hard, thin-bedded siltstone, limestone,
etc. that have bed thicknesses (and bedding plane joints) at say
3 in. (76 mm). In retrospect, it might have been better to have
chosen some form of weighted average so that all core pieces
could contribute to the RQD count. At the present state of
usage, however, it seems best to retain the requisite length and
to note on the boring log the reason for the low RQD. Employing
larger diameter drill bits, best drilling techniques, and short
runs will reduce core breakage along the incipient bedding plane

joints and will lead to higher, and more realistic, RQD values.
Such would not apply, of course, to pre-existing bedding joints

formed in nature by stress-relief and weathering.

Shale, claystones, interbedded sedimentary rocks. Five

queries have been assigned to this category. There is some
overlap with the previous comments. The argillaceous rocks
present the most trouble because they are weak and susceptible to

breakage during drilling, handling, and storage due to vibration

and moisture changes. Routine drilling and logging will not
give good samples or correct RQD values. The more careful

48
technique mentioned in the previous section must be employed to
obtain optimal results.
The senior author recalls examining cores on a hydro project

in Columbia many years ago. The shale cores had broken into

disks (poker chips) and RQD logging appeared next to impossible.

However, when going to the drill site and examining the cores as

they were retrieved, the cores were seen to be intact across the

lightly bonded bedding planes and the cross-cutting joints could

easily be recognized. RQD logging was possible.

The applicability of the RQD to shales or claystones inter

bedded with hard limestone or sandstone was of common concern in

several of the inquiries. This is a common condition in many

parts of the United States as well as elsewhere. There is a

tendency for the harder core pieces to spin on the softer shale
and vice versa. Shorter runs, 2-1/2 ft. to 5 ft. (0.75 to 1.5 m),

and larger core diameter usually result in improvement.

It would be of value for the different Districts that have

this interbedded geology to conduct a series of field tests in
which adjacent holes are drilled with variations in core

diameter, run length, and drilling techniques to isolate the most

important variables.

Limestone with solution cavities. Four comments dealt with

this condition. The RQD should not be isolated from the site

geology. The presence of a cavity within a core run should be

duly recorded on the boring log. In addition to the -overall RQD,

partial "artificial" run lengths can be shown with the

appropriate RQD for each, including zero for the cavity.

49
 Volcanics and metamorphics. One comment dealt specifically
with basalts and metamorphics, noting that these strong rock
masses may receive lower ratings than they deserve because of the
elimination of short core pieces. The authors agree. Rock
masses that contain tight, interlocked, irregular discontinuous
joints may be quite strong, impervious, and of high modulus.
At a recent project in Argentina, a wide, highly fractured
zone in andesite between two small faults was questionable as a
foundation for two blocks of a high concrete gravity dam because

of its fractured nature (mostly small pieces in the muck pile

following blasting; highly fractured appearance in-situ; and RQD
of about 25 in several of the short borings that had been
recently made to investigate the zone). Seismic traverses were
performed and surprisingly high P-wave velocities were obtained,
around 13,000 to 15,000-ft/sec. (4,000 to 4,500-m/sec.), values
similar to those for the adjacent less fractured andesite that
contained hard rhyolite intrusions. A closer examination was
made of the highly fractured zone and it was noted to be very
tight, difficult to remove with a pick, and the joints were

rehealed with hard epidote coatings. The thin hard coatings were
sufficient to improve the rock mass quality and make it

acceptable foundation rock; the coring and the blasting, however,

had broken the bonding of the joints, resulting in small pieces.
This example is one of many, no doubt, where the RQD gives
values too low for the rock with respect to bearing capacity and

modulus. But, for production of aggregate or riprap one might

say the RQD gave values consistent for those uses.

50
Orientation effects (2D)
the bias
Three questions or comments were received regarding
with
in RQD that may result from differing borehole orientations
respect to joint orientation. The problem is not severe where 3
be some
or 4 joint sets exist, although, even then, there can
bias when the boreholes parallel one of the sets.
The major problem is created when there is a Predominant

joint set, such as foliation or schistosity joints in metamorphic

bedded
rocks, or one or two vertical joint sets in horizontally
sedimentary rocks. For best results from the viewpoint of good
joints
core recovery, less breakage, and crossing the predominant
normal
at their true spacing, the borehole orientation should be

to the joints. Such orientation is often not practical but an

should
intersection of no greater than 45 degrees to 55 degrees
be attempted.
Where the intersection is at a steep angle, say 60 degrees

to 90 degrees, in addition to the greater potential for core

A
breakage, there can be a considerable bias in one of two ways.
or only
borehole may miss the predominant jointing altogether
cross it once or twice, leading to a higher RQD. On the other
and follow
hand, the borehole may hit a joint from the beginning
and to
it for a considerable distance, leading to core breakage

no pieces of cylindrical core. So as not be penalize the rock

is
quality too greatly, the center-line or axis measurement
recommended, as previously discussed.
more than one
A recommended procedure that has been used on
set exists, is
occasion by the writers, where a predominant joint

51
to drill the boreholes at both favorable and unfavorable crossing
angles so as to determine the directional bias. Notes can be
added to the boring logs pointing out this fact.

Desirability of Additional Geological Observations (3)

The importance of observing and recording other rock mass

characteristics was commented on by eight inquirers. The authors

certainly agree that RQD does not stand alone when attempting to

describe or characterize rock mass behavior (Deere, 1963). The

RQD may be characterized as a simple index, analogous to the SPT
blow count for soils, that has not only many useful design
correlations but also many limitations.

Joint conditions (3A)

It was recognized in the development of the RQD that many of
the important joint characteristics would not be included in the
RQD procedure and that additional engineering geological
observations and description would be necessary. The later rock
mass classification systems of Bieniawski (1973) and Barton et
al., (1974) do include most of the important joint character

istics, in fact, an important contribution of their systems was

to provide check lists of joint characteristics to be determined
from the core logging and field mapping.

Local geologv, weathering, fracture freauency (3B)

The many excellent comments received on these topics speak
for themselves. The RQD is only one of several tools or

52
techniques that help in understanding the site geology, siting
structures, and selecting foundation depths or tunnel supports.
It should not be used without a good knowledge of the
lithology, stratigraphy, and
local geology including weathering,

structural features.

Applications to EngineerinQ and construction (4)

General (4A)
One
Two comments or queries were assigned to this category.

asks for revision and expansion of the RQD - Rock Quality Table
to have, ,'... built in restrictions to prevent misinterpretation
for
of the rock quality descriptions (very poor - excellent)
rock
qualifying the meaning of the terms as applied to different

types and to the design of various types of structures,

tunneling, excavations and foundations."

The suggestion is good but because the two new classifi

and of Barton (1988) have
cation systems of Bieniawski (1988)

improved on the RQD and have more recent and more comprehensive
case histories relating to various design and construction
experience, it appears advisable to use their relationships.
a design
The second comment relates to the use of the RQD as
Certainly, the
aid without the consideration of other factors.
be
site geology with all of its pertinent factors must
of
considered. As noted above, the newer classification systems

Bieniawski and of Barton are recommended. They include the RQD

factors.
but as only one of several other important

53
 The RQD in itself should not be modified, in the authors
opinion; the usefulness of the RQD is in its simplicity. The low
RQD values act as a "red flag" to the engineering geologist and
rock engineer who must investigate the- cause of the low values -
rock weathering, shear zone, thin bedding, etc., or poor drilling
techniques. The RQD is not an end in itself but an indicator of
conditions to be investigated and explained.

Excavation. dredging. underwater blasting (4B)

Three queries were received on the general subject of
excavation. The RQD can not stand alone as a correlation tool
with excavatability, but has been used as one of the parameters
in excavatability prediction. The junior author has made his own
successful predictions of excavatability on projects using: (1)
RQD measured on short intervals and unconfined compressive
strength for a prediction of excavatability of a slurry wall with

a clamshell, and (2) the RQD, unconfined compressive strength,

and seismic refraction velocity for prediction of single-tooth

rippability with a D-8 dozer, all correlated by field rippability

tests.
Recently, two excavatability prediction systems using RQD as
a parameter have been published. Correlations have been
developed using several case histories. Smith (1986) utilizes
the RMR System to estimate rippability. Kirsten (1988)
characterizes excavatability for trenching, digging, dozing, and
ripping using a modified Q System.

54
 For any type of underwater excavation, it is important to

have a sufficient number of well controlled borings so that a

realistic geologic profile can be prepared with appropriate

descriptions and parameters for each geotechnical unit.

Foundations, in-situ modulus (4C)

Approximately one-third of the respondents queried the

usefulness of the foundation bearing or in-situ modulus

correlations with RQD. These topics are described within Part

IV.

Both correlations are useful as starting points and should

be utilized in conjunction with other correlations or as checks

with field tests. Kulhawy's (1978) model for rock foundation

settlement and Bieniawski's (1978) correlation with deformation

modulus both improve on the RQD concept by including joint

properties.

The senior author's primary use of RQD for foundations is

for project siting when comparing depths of excavation to
acceptable rock for high concrete dams.

Tunnels (4D)

In response to a query on the applicability of RQD to

openings at great depths, the authors believe it is still

applicable for its primary use for project siting or "red

flagging"- of zones of poor quality rock. Core disking due to

high in-situ stress may occur which would preclude the use of the

RQD. The ratio of in-situ stress to the intact unconfined

55
compressive strength controls the core disking as it would tunnel

wall stability. Reference should be made to Barton (1988) for

more on this topic.

Horizontal borings along the tunnel line would be most
helpful in intersecting the steep structural features of shears,

faults, and closely jointed zones and allowing the RQD to be

determined for each. During the construction of several tunnels

horizontal "feeler" or probe holes have been drilled from the

face of the tunnel to give advance warning of weak zones and any
contained groundwater.
Occasionally, horizontal holes from the tunnel portal area

have been drilled during the exploratory phase. For practical

purposes, however, most of the exploratory drilling for tunnels

will be vertical with some angled holes to cross suspected weak

fractures and to give a 3-dimensional picture of the bedding and

the jointing.

Erosion resistance, rouahness coefficient (4E)

Two interesting questions were received regarding erosion of

rock masses caused by flowing water in hydraulic tunnels and in

channels. The senior author has inspected numerous unlined

diversion tunnels, pressure tunnels, trailrace tunnels, and rock

channels after a few months to a few years of operation.

Pressure tunnels have low flow velocities of perhaps 13 ft.

to 16 ft. (4 m to 5 m) per second maximum and fair to good

quality rock has resisted erosion very well. In these cases the

zones of weak and heavily fractured rock and shear zones had been

56
protected by concrete or by reinforced shotcrete and rock bolts.
Diversion tunnels during flood will be subjected to higher
velocities, perhaps 36 ft. to 46 ft. (11 m to 14 m) per second,

and some erosion has been noted in both the invert and lower side

walls in unprotected weak zones and heavily fractured zones.

Similar velocities and even higher may occur in spillway

channels; weak zones have been eroded considerably (schistose

zones, within a more massive gneiss in one case).

While the RQD can be helpful in detecting the presence of

the weak zones and in delineating the more massive rock areas, it

probably can do no more than could good engineering geology

descriptions in predicting erodibility or roughness.

General Usefulness of ROD (5)

Many comments were received on both the favorable

experiences and the shortcomings of the RQD. This discussion

summarizes the comments received and attempts to place the use of

the RQD in perspective.

Favorable experience (SA)

Three comments were singled out from the responses that

specifically indicated the RQD has been helpful. One comment

noted that the index had been found to be a practical parameter

for estimating rock core quality and no problems had been
experienced with its application as an engineering index.

Another, while noting its shortcomings, felt that it had allowed

coordination of the nature of the rock mass to engineering

57
characteristics in a quick and simple manner -- and any

modification that would detract from its simplicity would be a

disservice. And, further, "...What the RQD system does is add

the experience factor for the inexperienced people." A third

comment noted that the RQD was one tool available, like other

index properties, for the evaluation of rock behavior in various

engineering applications.
The authors believe that the usefulness of the RQD can be
divided into these basic categories:
o The "Red FlaQ" effect. The RQD directs the attention
of the engineering geologist and design engineer to
areas of rock with poor engineering properties. These

are areas that may control the design of a project and

should be avoided or have sufficient means available to
cope with them.

o DesiQn guidance. Correlations of RQD with rock

properties and project performance provide preliminary

design guidance for structures on rock.
o Stimulation of Profession. The RQD was developed at a

time when the field of Rock Mechanics was in its

infancy. RQD helped focus the attention on the

importance of rock weathering and discontinuities and
on obtaining information from rock cores regarding

them. The RQD concept stimulated others, no doubt, to

"related studies of fracture frequency, scan-line

surveys, and to the development of modern engineering

classifications systems.

58
Shortcominas. limitations (5B)

The shortcomings and limitations of the RQD were popular

subjects and 10 comments or queries specifically addressed this

(See Appendix). The authors appreciate the many thoughtful

comments that were presented in the letters and are in general

agreement with the majority of them.
Perhaps the most common compliant was not against the RQD

p•r se but the manner in which it is often used in design as the

sole parameter without considering the geologic details and the

overall geologic evaluation of the site. Certainly, the core

logging should be done at the site by a qualified engineering

geologist or geotechnical engineer at the time of the drilling
and not left to the driller or technician, or done in the

laboratory days or weeks later after the core has been

transported, dried, stress-relieved, and otherwise disturbed.

The structural, hydraulic, or highway design engineer could

well misuse the correlation tables without the input of the

engineering geologist, geotechnical engineer, or rock mechanics

specialist who has knowledge not only of the critical geologic

details but also of the precedent in engineering design and

construction. The RQD can not stand alone. Its inclusion into

the later classification systems that include other geological

factors -(Barton, 1988; Bieniawski, 1988) was a logical

progression in use.

One- comment noted that the RQD was not very helpful in

selecting foundation depth in weathered rock. This experience is

contrary to the authors' experiences where such application has

59
had excellent results (Deere and Deere, 1988), not only for
foundation depth but also for tunneling and selection of dam

excavation depths.
A few comments dealt with the simplicity of the method,
which is both favorable and unfavorable, and the misuse that may
result from the qualitative terminology of good, fair, etc., that
may not always apply to the specific site or a specific
engineering problem. The authors agree and recommend that the
more comprehensive classification systems noted above be applied
-- as they are developing a good base of case histories.

60
 PART VII: CONCLUSIONS

Twenty years' experience with the RQD logging and

in
application of the RQD index to engineering has been discussed

the previous parts of this report. The main conclusions may be

summarized as follows:
1. Core diameters for RQD logging should normally be of NQ
or NWX (NWM) size; for weak argillaceous or foliated

rocks, larger sizes are preferred; and the smaller BQ

and BWX sizes should be discouraged and; when used,

should be identified with a disclaimer.

2. Length measurements of the core pieces should be made

along the centerline (axis) as advocated by the ISRM
(1978, 1981); core breaks caused by drilling and

handling should be fitted together and counted as one

piece; and the requisite length should be retained as
4-in. (100 mm) for all size cores.

Fresh and slightlv weathered rock should be used in the

3.
RQD count; moderately weathered rock that resists hand
breakage should be included but an asterisk used with
the RQD (RQD*); and highly weathered rock (that breaks
under hand pressure), completely weathered, and

residual soil should not be included.

Length of coring run ideally should be 5 ft. (1.5 m)

4.
- but realistically may be longer or shorter; for good
quality rocks which give 100 percent core recovery,

10-ft. (3-m) runs are acceptable; for difficult

61
 schistose, laminated, soft and hard interbedded, and
rocks with unfavorable joint or bedding orientations,
short run lengths of 2-1/2 ft. to 5 ft. (0.75 to 1.5 m)
or less are recommended; short "artificial" run
lengths, or intervals, may be created when logging the
core to identify zones of vastly different RQD.
5. ROD correlations with tunnel support requirements, in
situ modulus, allowable bearing pressure, and fracture
frequency are available in the literature, some of
which have been included herein; these are still
considered helpful in preliminary studies. Perhaps the
most important use of the RQD in practice is in early
delineation or "red flagging" of zones of poor rock.
6. More recent classification systems (Bieniawski RMR,
Barton Q) have included the RQD together with other
parameters that broaden the scope and more closely
define the rock quality for engineering purposes; these
have gained international acceptance and are
recommended herein.
7. Pertinent auestions and comments regarding the RQD
logging procedures and utilization within the various
Districts and Divisions of the US Army Corps of
Engineers have been reviewed and discussed under five
categories; many of the concerns were the same as those
rioted in the first six conclusions herein; of
particular concern were the additional topics: first,
the need for drilling supervision and prompt logging in

62
 the field by a qualified engineering geologist or
geotechnical engineer of cores obtained by the best
drilling equipment and techniques; second, the possible

misuse of RQD in design by using it as the sole

parameter without the necessary geological and geotech

nical input; and third, the difficulties of obtaining

intact cores and reliable RQD values in shales and in

interbedded hard and soft rocks.

8. A research Program is recommended wherein the Corps of

Engineers does comparative studies in bedded soft and

hard rocks of recoveries and RQD's in adjacent borings

drilled at differing angles, core sizes, lengths of

drill run, and drilling techniques.

The authors acknowledge the interest, questions, and

comments of the users of the RQD in the various Districts and

in
Divisions of the US Army Corps of Engineers which have helped
focusing on critical issues.

63
 REFERENCES
Barton, N., Lien, R., and Lunde, J., (1974), "Engineering Class
ification of Rock Masses for the Design of Tunnel Support,"
Rock Mechanics, Vol. 6, pp. 189-236.
Barton, N., (1988), "Rock Mass Classification and Tunnel
Reinforcement Selection Using the Q-System," Roc
Classification Systems for Enqineering Purposes, ASTM STP
984, ed. Louis Kirkaldie, American Society for Testing and
Materials, Philadelphia, pp. 59-88.
Bieniawski, Z.T., (1973), "Engineering Classification of Jointed
Rock Masses," Transactions. South African Institution of
Civil EnQineers, Vol. 15, pp. 335-344.

., (1974), "Geomechanics Classification of Rock

Masses and Its Application in Tunneling," Proceedincs. 3rd
International Congress on Rock Mechanics, ISRM, Denver,
Natl. Acad. Sci., Washington, Vol. 2A, pp. 27-32.
., (1976), "Rock Mass Classification in Rock
Engineering," Proc. Symp. on Exploration for Rock
Engineering, ed. Z.T. Bieniawski, Balkema, Rotterdam, pp.
97-106.

., (1978), "Determining Rock Mass Deformability;

Experience from Case Histories," International J. Rock Mech.
Min. Sci., Vol. 15, pp. 237-247.

., (1988), "The Rock Mass Rating (RMR) System

(Geomechanics Classification) in Engineering Practice," Rock
Classification Systems for Engineering Purposes, ASTM STP
984, ed. Louis Kirkaldie, American Society for Testing and
Materials, Philadelphia, pp. 17-34.
Cecil III, O.S., (1970), Correlations of Rock Bolt--Shotcrete
Support and Rock Quality Parameters in Scandinavian Tunnels,
Ph.D. Thesis, University of Illinois, Urbana, 414 pp.
Coon, R.F. and Merritt, A.H., (1970), "Predicting In-Situ Modulus
of Deformation Using Rock Quality Indexes," Determination of
the In-Situ Modulus of Deformation of Rock, ASTM STP 477,
American Society for Testing and Materials, Philadelphia,
pp. 154-173.
Cording, E.J., and Deere, D.U., (1972), "Rock Tunnel Supports and
Field Measurements," in Proceedings, North American Rapid
ExcaIation and Tunneling Conference, Chicago, Society of
Mining Engieers of AIME, New York, Vol. 1, pp. 601-622.
Deere, D.U., (1963), "Technical Description of Rock Cores for
Engineering Purposes," Felsmechanik und InQenieurgeologie
(Rock Mechanics and Engineering Geology), 1:1, pp. 16-22.

64
 01P (1968), "Geologic Considerations", Chapt. 1, Rock
Mechanics in Engineerinq Practice, ed. K. Stagg and 0.
Zienkiewicz, Wiley, N.Y., pp. 1-20.

Deere, D.U., Hendron, A.J., Jr., Patton, F.D., and Cording, E.J.,
(1967), "Design of Surface and Near-Surface Construction in
Rock:, Failure and Breakaqe of Rock, ed. C. Fairhurst, Soc.
of Min. Eng., AIME, N.Y., pp. 237-302.

Deere, D.U., Peck, R.B., Monsees, J.E. and Schmidt, B., (1969a),
"Design of Tunnel Liners and Support Systems," UIUC Final
Rept. for U.S. Department of Transportation (OHSGT),
Contract 3-0152, 287 pp. (available from the U.S. Department
of Commerce, NTIS, Springfield Virginia, Publ. No. PB 183
799).

Deere, D.U., Merritt, A.H., Coon, R.F., (1969b), "Engineering

Classification of In-Situ Rock," Tech. ReDt. No. AFWL-TR-67
144, Kirtland Air Force Base, New Mexico, 280 pp.
(available
from the U.S. Department of Commerce, NTIS, Springfield
Virginia, Publ. No. AD 848 798).

Deere, D.U., Peck, R.B., Parker, H.W., Monsees, J.E., and

Schmidt, B., (1970), "Design of Tunnel Support Systems,"
Highwav Research Record, No. 339, Highway Research Board,
pp. 26-33.

Deere, D.U., and Deere, D.W., (1988), "The Rock Quality

Designation (RQD) Index in Practice," Rock Classification
Systems for EnaineerinQ Purposes, ASTM STP 984, ed. Louis
Kirkaldie, American Society for Testing and Materials,
Philadelphia, pp. 91-101.

Fookes, P.G., and Horswill, P., (1970), "Discussion, Session A:

Properties of Rocks; Foundations of Surface Structures,"
Proc. Conf. on In-Situ InvestiQations in Soil and Rocks, The
British Geot. Soc., London, pp. 53-57.

Goodman, R.E., (1980), Introduction to Rock Mechanics, John Wiley

and Sons, New York, p.478.

Goodman, R.E., and Smith, H.R., (1980), "RQD and Fracture

Spacing," Journal of the Geotechnical Engineering Division,
ASCE 106:191-193.

Hendron, Jr., A.J., (1968), "Mechanical Properties of Rock,"

Chapt. 2, Rock Mechanics in Engineering Practice, ed. K.
Stagg and 0. Zieniewicz, Wiley, N.Y., pp. 21-53.

Heuze, F.E., (1971), "Sources of Errors in Rock Mechanics Field

Measurements and Related Solutions," Int. J. Rock Mech. Min.
Sci., Vol. 8, pp. 297-310.

65
Hoek, E., and Brown, E.T., (1980) Underground Excavations in
Rock, Institute of Mining and Metallurgy, London.
Hudson, J.A., and Priest, S.D., (1983), "Discontinuity Frequency
in Rock Masses," International Journal of Rock Mechanics and
Mining Sciences, 20:73-89.
International Society for Rock Mechanics, (1978), "Suggested
Methods for the Quantitive Description of Discontinuities in
Rock Masses," Commission on Standardization of Laboratory
and Field Tests, International Society for Rock Mechanics,
Int. J. Rock Mech. Min. Sci., Vol. 15, pp. 319-368.
, (1981), ISRM Sugaested
Methods: Rock Characterization. Testing and Monitorin , ed.
E.T. Brown, Pergamon Press, London, 211 pp.
Kazi, A., and Sen, Z., (1985), "Volumetric RQD: An Index of Rock
Quality," Proceedinas of the International Syposium on
Fundamentals Rock Joints, BJorklliden.

Kirsten, H.A.D., (1988), "Case Histories of Groundmass Character

ization for Excavatability," Rock Classification Systems for
Engineerina Purposes, ASTM STP 984, ed. Louis Kirkaldie,
American Society for Testing and Materials, Philadephia, pp.
102-120.

Kulhawy, F.H., (1978), "Geomechanical Model for Rock Foundation

Settlement," Journal of the Geotechnical Engineering
Division, ASCE, Vol. 104, No. GT2, pp. 211-228.

Little, A.L., (1969), "The Engineering Classification of Residual

Tropical Soils," Proc. 7th Int. Conf. Soil Mech. Found.
Engng., Mexico, Vol. 1, pp. 1-10.
Merritt, A.H., (1972), "Geologic Predictions for Underground
Excavations," Proc. North Amer. Rapid Excav. and Tunneling
Conf., 1:115-132.

Palmstrom, A., (1982), "The Volumetric Joint Count: A Useful and

Simple Measure of the Degree of Rock Mass Jointing,"
Proceedings of the 4th Congress of the International
Association of Engineering Geoloav, 5:221-228.
Peck, R.B., (1976), "Rock Foundations for Structures," Rock
Engineering for Foundations and SloRes, Specialty Conference
of Geotech. Div., ASCE, Boulder, Vol. II, pp. 1-21.
Peck, R.B;, Deere, D.U., Monsees, J.E., Parker, H.W., and
Schmidt, B., (1969), "Some Design Considerations in the
Selection of Underground Support Systems," UIUC Final Rept.
for U.S. Department of Transportation (OHSGT) (UMTA), 108
pp. (available from U.S. Department of Commerce, NTIS,
Springfield, Virginia, Publ. No. PB 190, 443).

66
Peck, R.B., Hanson, W.E., and Thornburn, T.H., (1974),
"Foundations on Rock," Foundation Enaineering. 2nd ed., John
Wiley and Sons, Inc. New York, N.Y., pp. 361-371.

Priest, S.D., and Hudson., J.A., (1976), "Discontinuity Spacings

in Rock," International Journal of Rock Mechanics and MininQ
Sciences, 13:135-148.

Priest, S.D., and Hudson, J.A., (1981), "Estimation of

Discontinuity Spacing and Trace Length Using Scanline
Survey," International Journal of Rock Mechanics and Mining
Sciences, Vol. 18, pp. 183-197.

Sen, Z., (1984), "RQD Models and Fracture Spacing," Journal of

Geotechnical EnqineerinQ, ASCE 110:203-216.

Sen, Z.,and Kazi, A., (1984), "Discontinuity Spacing and RQD

Estimates from Finite Length Scanlines," International
Journal of Rock Mechanics and Mining Sciences, 21:203-212.

Serafim, J.L., and Pereira, J.P., (1983), "Considerations of the

Geomechanics Classification of Bieniawski," Proc. Int.
Symposium on Engineering Geology and Underground
Construction, LNEC, Lisbon, Portugal, Vol. 1, pp. II. 33
II. 42.

Smith, H.J., (1986), "Estimating Rippability by Rock Mass

Classification," Proc. 27th U.S. Symposium on Rock
Mechanics, AIME, New York, pp. 443-448.

Stagg, K.G., and Zienkiewicz, O.C., (1968), Rock Mechanics in

Engineering Practice, Wiley, N.Y., 442 pp.

U.S. Army Corps of Engineers, (1978), "Tunnels and Shafts in

Rock,"Encfineer Manual, EM 1110-2-2901.
Wallis, P.F., and King, M.S., (1980), "Discontinuity Spacings in
a Crystalline Rock," International Journal of Rock Mechanics
and Mining Sciences, Vol. 17, pp. 63-66.

67
 APPENDIX A

COMMENTS AND QUESTIONS ON THE RQD

FROM THE US ARMY CORPS OF ENGINEERS

A-i
 APPENDIX A
COMMENTS AND QUESTIONS ON THE RQD
FROM THE U.S. ARMY CORPS OF ENGINEERS

CONTENTS
1. MECHANICS OF DETERMINING ROD
IA. CORE DIAMETER
lB. LENGTH MEASUREMENT OF THE CORE PIECES

- Position of the Measurement

- Core Length of Twice the Diameter
- Distinguishing between Natural and Induced
Fractures
IC. LENGTH OF CORING RUN AND RQD INTERVAL
2. SPECIAL ROD LOGGING PROBLEMS
2A. DRILLING EQUIPMENT AND TECHNIQUES

2B. PROMPT LOGGING OF CORES

2C. APPLICABILITY TO VARIOUS ROCK TYPES

- Problems
- Shale, Claystone; Interbedded Sedimentary Rocks
- Limestone with Solution Cavities
- Volcanics and Metamorphics

2D. ORIENTATION EFFECTS

3. DESIRABILITY OF ADDITIONAL GEOLOGICAL OBSERVATIONS

3A. JOINT CONDITIONS

3B. LOCAL GEOLOGY, WEATHERING, FRACTURE FREQUENCY

4. APPLICATIONS TO ENGINEERING AND CONSTRUCTION

4A. GENERAL

4B. EXCAVATION, DREDGING, UNDERWATER BLASTING

4C. 7OUNDATIONS, IN-SITU MODULUS

4D. TUNNELS
4E. EROSION RESISTANCE, ROUGHNESS COEFFICIENT

A-2
5. GENERAL USEFULNESS OF ROD

5A. FAVORABLE EXPERIENCE

5B. SHORTCOMINGS, LIMITATIONS

A-3
 1. MECHANICS OF DETERMINING ROD

IA. CORE DIAMETER

Ouestion
1A 1: RQD was based upon use of NX, double tube, core
sampling which has not been used by Mobile District for
a number of years. Has any correlative work been done
between NX, NQ wire line and/or other sizes and types
of sampling?
1A 2: Is it valid to apply RQD to cores greater than NX-size
by increasing the length of core used in determining
the Modified Core Recovery to twice the core diameter
(i.e., count only 12-inch long pieces for 6-inch
diameter core)?
1A 3: Are RQD values applicable to other than NX size core?
1A 4: RQD is based on NX core - how can RQD be used,
measured, correlated on larger or smaller diameter
cores?
1A 5: I would like to see the RQD table revised by including
information on applying this system to different rock
lithology, in-situ geology, core size, etc...

1A 6: ... It does not provide for core diameters larger or

smaller than NX and does not give an indication of
highly fractured zones within core runs...

1A 7: ... In recent years, this District has utilized 4" core

barrels almost exclusively though we are now making
considerable use of HQ wire line equipment. Due to the
nature of our rocks, highly fractured in Puerto Rico
and weak limestones in Florida, we do not use NX size
core equipment. When the question arose among field
geologists as how to measure fragment length in core
other than NX, we decided to piece the core run
together and use the top center of core as a reference
line. It was apparent we could not measure the long or
short side of individual pieces and have a direct
-correlation between 4" and NX core. We decided a
center reference line would produce a medium length
width would be representative regardless of the core
diameter...

1A 8: -The "Geotechnical Handbook" being prepared-should

include a discussion comparing RQD values taken on
similar rock, but cored with a different size or type
of core barrel. I have been told that Dr. Deere
intended RQD's to be used with only NX size cores.
Several texts discuss the determining of RQD values on
A-4 APPENDIX
 core 50 mm and larger. As you know a larger diameter
core results in better core recovery and in most cases
a higher RQD values.

1A 9: Can RQD be reasonably extended to core sizes other than

NX? If so, which ones and why, or why not?

1A 10: Is there any experience of relating RQD to diameter of

the core. It is our understanding that RQD was
developed for NX core and length of solid core divided
by the core diameter equal to 2. With 3-inch core this
would result in RQD based on solid pieces being 6
inches rather than 4 inches as in NX core. It is
believed that common practice is to using four inch
solid pieces in all core sizes. This should be
clarified.

1A 11: RQD is of extremely limited use in today's core

drilling work when NX is only one of the choices for
processing the work in the most cost-efficient and/or
highest rock recovery/quality manner. More and more we
are using wire line systems. Will this re-study
address correlations of RQD to other sizes of core
besides NX? In particular, correlation should be made
for large diameter (4", 2-3/4") as well as wire line
sizes such as PQ, HQ, etc. and triple tube c.b.

1A 12: One of the major drawbacks of RQD is its relationship

to core size, i.e., the larger the core diameter, the
smaller the RQD. To be useful as a true index property
of a rock mass, RQD should be independent of hole size.
Can compensation be made for this problem?

1A 13: ... We have had much discussion recently about the

definition and use of RQD, and we would like to relay
our feelings on the following three areas: (1) As we
understand it, RQD is calculated counting only those
pieces of drill core greater than 4 inches in length.
This does not seem to be the appropriate method since
it is not a true reflection of the condition of the
rock mass in place; (2) the use of RQD by the Corps of
Engineers is confined only to NX core. This is very
restrictive since a major portion of the drill holes in
the Northwest are larger diameter (HQ) than NX...

lB. LENGTH MEASUREMENT OF THE CORE PIECES

ouestion

Position of the Measurement:

lB 1: Measuring RQD needs clarification - where on core?

A-5 APPENDIX
lB 2: ... When the question arose among field geologists as to
how to measure fragment length in core other than NX,
we decided to piece the core run together and use the
top center of core as a reference line. It was
apparent we could not measure the long or short side of
individual pieces and have a direct correlation between
4"1 and NX core. We decided a center reference line
would produce a medium length which would be represent
ative regardless of the core diameter...

lB 3: Fractures parallel to the axis of coring need to be

accounted for. TVA uses the rule that if a section of
core is split into two longitudinal halves of length
"x" that is more than 4" in length, the value of x/2 is
counted in the summation of intact pieces of core.
Core Length of Twice the Diameter:
lB 4: Is it valid to apply RQD to cores greater than NX-size
by increasing the length of core used in determining
the Modified Core Recovery to twice the core diameter
(i.e., count only 12-inch long pieces for 6-inch
diameter core)?
1B 5: Is there any experience of relating RQD to diameter of
the core? It is our understanding that RQD was
developed for NX core and length of solid core divided
by the core diameter equal to 2. With 3-inch core this
would result in RQD based on solid pieces being 6
inches rather than 4 inches as in NX core. It is
believed that common practice is to use four solid
pieces in all core sizes. This should be clarified.

1B 6: Based on the original data available, the 4-inch core

length was adopted for RQD determinations. Does the
data from the last 20 years support this, or would some
other length be better?
lB 7: ... If the RQD of the core is measured using only
natural breaks, it should be close to the value
obtained using downhole photographic methods... Many
of the drill holes presently drilled here in the
Northwest are HQ-sized which, by the Corps of
Engineers' policy, cannot be used in the calculation of
RQD. If the above method of calculating RQD were to be
used, there would be no difference in the results of
calculations made in NX holes or 6-inch holes since
they are both a measure of what is in the ground. The
-present method of RQD calculation would probably arrive
at different values because of differing mechanical
stresses on the drill core during drilling causing

differing degrees of breaks along healed fractures,

bedding planes, or foliations. Again, counting only
A-6 APPENDIX
 those naturally occurring breaks allows any size of
drill hole to be used in the calculation of RQD.

Distinguishing Between Natural and Induced Fractures:

lB 8: Clarification on natural vs. mechanical break needed.

lB 9: Are there techniques other than fitting core pieces,

which can be used by a field geologist or technician,
to differentiate between natural (in-situ) separations
and drilling induced separations in shales or shaley is
siltstone, sandstone sequences? This determination
vital to an accurate definition of the RQD.
Stress
lB 10: Stress relief discing of deep cores does occur. cores
relief discing is a larger problem with smaller
that with larger cores. Stress relief discing should
no be included in RQD counts of core breaks. Some
geologists could confuse discing as joints. Some
guidance on field logging techniques would be helpful.

lB 11: The definition of RQD, according to Technical Report

GL-85-3, Geotechnical Descriptions of Rock and Rock of
Masses, is a method used to describe the condition
the rock mass from core borings. It is assumed that
the condition of the rock mass refers to those it.
naturally occurring breaks and fractures within
This is not what is described when the RQD is greater
calculated counting only those pieces of core
than 4 inches in length. In describing the rock mass
in-situ, one has to look at only the spacing between by
the unhealed fractures in the rock mass intersected
the drill hole. If the core could be obtained without
places
the mechanical forces that drilling and handlingand
on the core, the only breaks would be the open
closed fractures and shear zones for faults. None of
be broken and bedding of
the healed fractures would during
foliation parting would not take place as occurs
The RQD would then be the
normal drilling operations. with
measure of all the cores containing natural breaks in
spacings of 4 inches or greater. This is a concern
the Walla Walla District because an RQD measured
looking at the core differs greatly from a RQD measured
or the borehole
downhole with the NX borehole camera,
of the core is measured
analysis package. If the RQD
to the
using only natural breaks, it should be close
value obtained using downhole photographic methods.
mainly
lB 12: .In the New York District rock drilling is shallow
limited to the first 5 feet encountered during
borings. As a geologist who subsequently examines rock
of
core, I have found that general rock descriptions,
which RQD's are not always given, can often under for
estimate the rock quality, due to the difficulty
A-7 APPENDIX
 inspectors in distinguishing between natural and
mechanical breaks. The Districts can be faced with
claims for excavating equipment not being able to
remove "highly fractured" rock. Consequently I would
appreciate it if Dr. Deere would consider this problem
in his evaluation of RQD and possible ways to avoid it.

I am familiar with the different rock coring methods

and ways that core can be broken mechanically. I at
times have trouble identifying natural breaks but can
usually eliminate 3/4 of the breaks as mechanical.
Besides odd fractures I do not include tight fractures
which were opened by the coring process.
I have recently completed a 1-1/2 year deep-hole coring
program along a proposed flood diversion tunnel for the
Passaic River Basin in New Jersey. It is the first
time in recent years that extensive rock coring has
been done in the New York Districts. Rocks encountered
were shale and sandstone with 100-200 bedding joints
and basalt with 10o-200 stress release and 800-900
columnar joints. I took RQD values for only the 3
tunnel diameter zone which was usually at least 100
feet below the rock surface. The rock quality was good
to excellent though the cores themselves were highly
fractured. Fortunately because of the rock quality,
after eliminating mechanical breaks, I estimated the
RQD for most cases without measuring the rock core.
One day the consultants who were going to get the rock
data came out to inspect the drilling. They were
shocked to look at a highly fractured 10 foot core and
hear me give it a 95% RQD value. Another case of not
being able to detect the mechanical breaks. To satisfy
myself and reassure the consultants I made some graphs
for the GDM report using my drilling logs and downhole
camera photographs of undisturbed rock. Bill Tanner's
camera from the Southwestern Division lab was used
mainly to determine structures yet individual natural
fractures were also detected on the film.
Enclosed are copies of my graphs. I first totalled the
joints from both data sources and divided them into 100
intervals. As you can see I missed many natural breaks
S400). Part of the large discrepancy can be attributed
to my not logging each joint from a highly fractured
(ex. 2 joints per inch) zone on my drilling logs. The
percentage of missing joints can not be determined.

The enclosed graph comparing the undisturbed rock data

(photography) and the disturbed rock data (core logs)
show that the addition of missing natural breaks does
not reduce the RQD values. Instead, for the most part,
the undisturbed rock had higher RQD values than I
determined. The missing natural breaks add an average

A-8 APPENDIX
 of 1 joint per 11 feet which does not reduce RQD values
based on RQD measuring requirements eliminating rock
core less than 4 inches.

IC. LENGTH OF CORING RUN AND OF RQD INTERVAL

ouestion

iC 1: Only one question really comes to mind, did Dr. Deere

have any set dimensions such as core run, footage of
core box or total core hole footage for computation of
the index? For convenience and uniformity of
intervals, we used the total footage of a core box to
determine RQD.

1C 2: Clarification needed if RQD should be measured as a %

over every 5 feet, 10 feet, length of run, core box,
etc.

1C 3: No guidelines are uniformly followed in selecting

intervals for reporting RQD. For example, arbitrary
and varying lengths of core run up to 10 feet may be
given an average RQD. An average RQD of 50 over 10
feet does not satisfactorily reflect field conditions
where to upper 5 feet of rock has an RQD of 0 and the
lower 5 feet has an RQD of 100. Where abrupt changes
occur in rock quality, RQD values should be reported
separately for each interval.

1C 4: Since RQD is based on the total length of rock drilled,

the results are affected by the quality of not only the
rock but also the drilling process. Inappropriate bit
types, feed rates, water pressures, barrel adjustments,
core size and other factors can greatly affect the
percent rock recovered and it's condition. These
influences are particularly problematic in shales . . .
RQD could be based on the length of core recovered;
however, the wealth of experience and "feel" for ranges
would be jeopardized by such a change.

1C 5: ... We follow the guidelines in the South Atlantic by

Division Geotechnical Manual which is as described
Dr. Deere in his original paper. We divided the total
-length of sound, fresh pieces of rock over 4" long The
recovered during the run by the length of the run. the
contractor's position was that you should only use
length of the run in rock not total run. Our position
was, and still is, that unless you get 100% recovery
-(uncorrected) you have no way to know what-was "soil"
or what was "rock". I use the quotes because in our
case it was not a soil in the true meaning, but rather
a soft zone or layer of unconsolidated material within
harder, consolidated layers. Using the contractor's
technique, losses experienced were considered "soil"
A-9 APPENDIX
and all soft zones were also called "soil". In effect
you always get 100% recovery (uncorrected) in "rock".

Your RQD's are then computed based on this total

length, counting all pieces 4" or longer in length. It
naturally results in a much higher RQD.

A-10 APPENDIX
 2. SPECIAL ROD LOGGING PROBLEMS

2A. DRILLING EQUIPMENT AND TECHNIQUES

ouestion
2A 1: A factor that seems to be ignored in the application of
RQD is the skill of the drill operator. This seems to
be an important factor. How should it be evaluated?
2A 2: Since no two drillers, rigs or equipment will produce
the same results in sampling identical materials, what
"Mickey Mouse" factor(s) are to be applied to provide
comparable data? . . . The bit type, "stone" size and
distribution, rotational speed, tool weight, drilling
fluid type, pressure and volume, core barrel length,
use of drill collars and or "trash baskets" and length
of core runs are also factors which contribute to
recovery and condition of samples.
2A 3: Since RQD is based on the total length of rock drilled,
the results are affected by the quality of not only the
rock but also the drilling process. Inappropriate bit
types, feed rates, water pressure, barrel adjustments,
core size and other factors can greatly affect the
percent rock recovered and it's condition. These
influences are particularly problematic in shales.
2B. PROMPT LOGGING OF CORES
Ouestion
2B 1: If the sample condition is not observed upon removal
from the core barrel, during handling and boxing and
immediately logged by a qualified person, the data
presented may be far from the original characteristics.
We often have core samples which are not logged until
after transportation and sometimes days or weeks after
obtained.
2B 2: In the Huntington District, we work almost exclusively
in thin-bedded sedimentary shales and sandstones. We
have found, particularly in shales, that RQD becomes
-dependent on drilling techniques, core handling and
rapid deterioration due to fissility. We feel that RQD
is, at best, only a vary general indicator of quality
for shales, indurated clays and poorly consolidated
claystones.

A-11 APPENDIX
2C. APPLICABILITY TO VARIOUS ROCK TYPES

Question
General Problems
2C 1: ... I would like to see the RQD table revised by
including information on applying this system to
different rock lithology, in-situ geology, core size,
etc...

2C 2: Has sufficient experience developed to indicate that

RQD is more or less applicable to different rock
categories - igneous vs. sedimentary vs. metamorphic
or even within a given category, i.e., shale vs.
sandstone vs. limestone?
2C 3: The following would all tend to reduce the RQD:
foliated zones, fault zones, shale, cavernous
limestone, and thinly bedded competent rocks such as
limestone. Although the rock units above and below the
above-mentioned zones may be of substantial strength
and competence, the RQD would appear low. Would it be
possible to increase the RQD based on each in-situ
case?

2C 4: RQD does not distinguish between fractures, broken

rock, and thin interbedded formations with minimal
fracturing and weathering if the beds are less than
4 in. thick.

2C 5: Should there be a way to account for in influence of

features such as bentonite seams in shale or micaceous
layers in igneous rocks, which are intact but represent
definite planes of greatly reduced strength, using RQD.
2C 6: Several problems have been encountered in the field
application of RQD during core logging:

a. Difficult to use in soft rock formations where

drilling action causes separation along incipient
fracture planes.

b. Weak, brecciated-type rocks often don't break

along natural fractures, and therefore, rate high
in RQD when in reality they are often very poor in
rock mass quality.

Shale- Claystones, Interbedded Sedimentary Rocks

2C 7: Are RQD values applicable for soft rock cores such as

compaction type clay shales or in interbedded clay
shale and harder limestone?

A-12 APPENDIX
2C 8: The Vicksburg District has not used the RQD method and
does not anticipate having to work in an area where it
can be utilized anytime in the near future. The
Vicksburg District has worked with RQD information
gathered by other districts and was of the opinion that
it was inappropriate due to the soft nature of the
rock.

2C 9: Very little documentation is available regarding RQD,

application in various rock types. For example, thin
bedded limestone or even massive soft shale may have
similarly poor RQD values or opposite RQD values
depending upon when the shale is logged or who logs it.
QUESTION: Can more specific guidance be provided for
this aspect?

2C 10: The original correlations did not include weak rocks

such as some sandstones and shales. Can correlations
for these now be developed?

2C 11: Irrespective of the potential misuse of RQD, the

procedure itself has some inherent problems. For
certain types of lithology, the use of RQD can result
in a gross misunderstanding of the engineering
properties of the rock in-situ. Particularly
troublesome are the thin-bedded shaley limestones
common to much of Middle Tennessee, for example. Many
of these rocks, even though cored by experienced
drillers, tend to break along shaley laminations which
cannot be reliably designated as in-situ or drilling
induced to the extent that RQD for any length of core
run will be very low. The erroneous implication then
being that the rock is of poor engineering quality.
However, the presence of lamination in the bedrock does
not significantly influence the capability of those
rocks to sustain very high compressive loads. Owing to
the high degree of anisotropy in the mechanical
properties of these types of rock, reliance on the RQD
to any degree can be highly misleading in making an
engineering evaluation.

Limestone with Solution Cavities

2C 12: In addition, solution activity in carbonate rocks often

produces cavities and corresponding low RQD's and core
recoveries. However, if the cavities are isolated and
surrounded by hard continuous limestone, the low values
may misrepresent the quality of the continuous rock for
-support of loads.
For these reasons, an alternate parameter such as
fracture frequency can be a better indicator of rock
quality than RQD.

A-13 APPENDIX
2C 13: Weathering features such as solutioning and open seams,
voids, mud seams, etc. can not be distinguished from
fracturing or thin bedding.
2C 14: Another problem with using RQD to describe the
engineering qualities of limestone or other soluble
bedrock is that the RQD procedure does not consider the
,impact of the thickness or location of cavities within
the bedrock mass on its structural integrity. For
instance, a ten-foot run of core that is essentially
sound except for a two-foot thick void in the middle of
the run would have an RQD of 80%, which is described as
"good." In this case, even though the majority of the
rock is hard and competent, the presence of a void
comprising twenty percent of the mass cannot be
overlooked. In fact, it is the nature and location of
the void, not the condition of the recovered core, that
would be the most important issue affecting the
engineering properties of that bedrock.
2C 15: In addition, solution activity in carbonate rocks often
produces cavities and corresponding low RQD's and core
recoveries. However, if the cavities are isolated and
surrounded by hard continuous limestone, the low values
may misrepresent the quality of the continuous rock for
support of loads.
For these reasons, an alternate parameter such as
fracture frequency can be a better indicator of rock
quality than RQD. RQD could be based on the length of
core recovered; however, the wealth of experience and
"feel" for ranges would be jeopardized by such a
change.
More practically, it may be valuable to simply note the
conditions which can affect RQD and suggest the use of
another parameter if conditions are suspect.
Volcanics and Metamorphics
2C 16: Problems have been inherent with the system since its
inception and prevent our wholehearted adoption.
First, the elimination of short core pieces from
-determining rock quality causes some strong rock
masses, such as, basalt and metamorphics to receive
lower ratings than they deserve. Good examples are the
local diced basalt units which stand in vertical cliffs
(even overhanging) due to the irregular nature of the
-fracture planes.

A-14 APPENDIX
2D. ORIENTATION EFFECTS

Ouestion
result in
Drill hole orientation can, and often does, holes
2D 1: that
considerable bias in the RQD values, i.e., a
parallel major fracture sets could indicate
misleading high RQD value. How can this shortcoming be
addressed?

2D 2: There is a difference in RQD based on the orientationIs

of joints and bedding in relation to the bore hole.the
there some correlation that can be used or should
procedure contain some warning.
2D 3: How can RQD measurements on core from vertical borings
hope to give an accurate prediction of the effects of
vertical and high angle joints on the engineering
properties of a rock mass?

A-15 APPENDIX
 3. DESIRABILITY OF ADDITIONAL GEOLOGICAL OBSERVATIONS

3A. JOINT CONDITIONS

ouestion
3A 1: Continuity, planarity, mineral alteration along joints
and shear planes are properties of rock discontinuities
that have primary affect on rock mass properties. RQD
frequently does not reflect these important features.
3A 2: There should be some way to connect RQD to fracture
roughness or infillings, such as clay, etc. How can
RQD reflect the significant contribution of fracture
nature toward overall rock mass quality?
3A 3: Will RQD systems, such as rock mass classifications, be
expanded - will joint conditions, joint sets, water,
etc. have an increased role?
3B. LOCAL GEOLOGY, WEATHERING, FRACTURE FREQUENCY
ouestion
3B 1: In regard to your inquiry about our use of RQD
measurements, we use it routinely in core drilling.
However, I do not think it is useful in our practice.
Generally, I am trying to estimate the average depth of
weathering into rock for foundation design and cost
estimating. That usually involves looking at each core
carefully and trying to pick the depth where any
significant rock weathering stops. RQD does not help
much in that regard.
3B 2: We have found that RQD, although a neat number to
calculate, is not always the best method to use in
evaluating the drill core. RQD is not meant to be a
stand alone method of evaluating core and is best used
in the calculation of geomechanical properties of rock
masses. For other situations a fracture frequency plot
of the core is more useful since one can visually see
at once where the fractured areas occur in the drill
hole. We also make a fracture frequency plot of the
healed fractures, since they are typically healed with
chlorites and tend to part when excavated, as a part of
the evaluation for possible quarry material.
3B 3: Anyone who uses RQD as a tool to understanding the
engineering properties of bedrock must also understand
the limitations inherent in such a simplistic approach
to assessing the engineering properties of bedrock.
When considered by itself, outside the context of the
local geology, i.e., the lithology of the rock, the
geologic structure of the bedrock and the potential

A-16 APPENDIX
 influence of bedrock weathering, RQD becomes arecognizes
meaningless number. Deere (1968)* certainly
for
the limitations of the RQD system and the necessity
considering the overall geology of a site when
with the
designing engineering projects that interact must be
bedrock. Therefore, to be properly used, RQD
considered as only one small part of the overall
geologic evaluation and cannot be used as the sole
of
basis for determining the engineering qualities
bedrock.
has been
3B 4: ... Most of my experience:with the use of RQD an
favorable but I have noticed through the years
and misinterpretation of the
increase in the misuse
system by engineers not trained in the geotechnical
field. As most of us know, the RQD method is one of
to
many tools which must be used with other factorsmass.
determine the suitability of the total rock
There is a growing number of engineers and architects
to the RQD
(structural and highway) that have locked on went into
table without regard to the many factors that that must
the system or the many geological conditions or
be considered when designing foundations, tunnels,
excavation slopes.
An increasing number of firms are using inexperienced
core drill inspectors who are not trained in good
in poor
descriptive logging techniques which results values.
rock descriptions and total dependence on RQD
to
I suspect the AE regards the RQD method as a panacea
the rising cost of detailed geotechnical investigations
and rock testing.
many of
3B 5: ... We do not require use of the RQD system, but We
our geologists use it because of certain benefits.
do not mind the use, as long as additional ininformation
their logs
which is not provided by the RQD is given a logging
or reports. Its benefits are simplicity as
charts
tool, universal fame and published correlation
such as,
containing engineering design parameters,
modulus, shotcrete thickness needed, etc.

the end of the

References in this appendix can be found at
main -text.

A-17 APPENDIX
 4. APPLICATIONS TO ENGINEERING AND CONSTRUCTION

4A. GENERAL

Question
4A 1: I would like to see the RQD table revised by including
information on applying this system to different rock
lithology, in-situ geology, core size, etc. The table
should have built in restrictions to prevent misinter
pretation of the rock quality descriptions (very poor
excellent) by qualifying the meaning of the terms as
applied to different rock types and to the design of
various types of structures, tunneling, excavations and
foundations.

4A 2: RQD is a tool that has been widely used. It can also

be misused by the fact that a lot of design aids have
been developed using RQD along with other information
on rock properties. With these aids some designers can
take RQD alone to use in design. A discussion on the
intended use and limitations should be included if a
report is prepared.
4B. EXCAVATION, DREDGING, UNDERWATER BLASTING
Ouestion
4B 1: I would appreciate it if you would include among your
questions to Dr. Deere whether he has had any
experience in correlating RQD values with the use of
particular kinds of capacities of excavating and
dredging equipment (backhoe, clamshell, dipper, dredge,
etc.) without blasting.
4B 2: May I add for underwater blasting. Please add it to my
previous submission.

4B 3: We would like Dr. Deere's thoughts on the utility of

RQD in determining rock excavatability for both
dredging projects and surface construction projects.
4C. FOUNDATIONS, IN-SITU MODULUS

Question
4C 1: Some propose the use of RQD to arrive at allowable
bearing values for foundations bearing in bedrock. Are
tuch methods realistic or valid?
4C 2: It is our understanding that RQD was originally
developed as a method for tunnel design evaluations.
Over the past several years, it has become a generic
guide to rock foundations quality. As such, RQD is now

A-18 APPENDIX
 being applied to all types of foundation design by
some. At the present time, it is assumed that the use
of the RQD value by the engineer/geologist is based on
experience together with other known parameters or when
correlated to allowable contact pressures as referenced
in Table 22.2 of Peck, Hanson and Thorburn, "Foundation
Engineering" and other rock quality indices. However,
when supplied to an unknowing or inexperienced engineer
or contractor who looks at the RQD recorded on a raw
boring log, the resulting interpretation can be
misleading. QUESTION: Can the method of describing
the rock core be modified or the quality designations
be better defined for specific uses?
4C 3: In summary, RQD serves a purpose and can be a useful
tool to describe certain properties of rock core.
However, its design application is severely limited.
The fact that RQD is correlated with such terms as
"excellent", "good" and "poor", and is sometimes even
correlated with allowable bearing capacities (Peck,
Hansen, Thornburn, 1974), affords much opportunity for
its misuse. Designers must not rely on RQD alone as a
basis for foundation design decisions. At best, it can
only serve as a tool of limited use in the assessment
of the engineering qualities of bedrock.
4C 4: The original correlations between RQD and rock modulus
or deformation ratio did not appear to be exceptionally
good. Does more data verify these correlations or do
they need revision?
4C 5: RQD is an index of in-situ rock quality, but the
information provided by it affords only a rough
qualitative measure of rock quality. Attempts have
been made to relate RQD with modulus of deformation of
rock mass, but the procedure ignores many factors which
control the deformation modulus. Therefore, a large
number of engineers consider the RQD method of
evaluating rock mass modulus unreliable. Is it safe to
use the deformation modulus determined by the RQD
method for stability analysis of structures on rock
foundations?

4C 6: The deformation modulus of rock mass not only depends

upon the number of discontinuities, but also on their
characteristics. RQD does not provide information as
to these characteristics; therefore, it is not
appropriate to correlate the RQD in its present form
,with deformation modulus of rock mass. Can the form of
the RQD be changed to include the characteristic of
discontinuities?

4C 7: To perform stability analysis of concrete structures on

rock foundations, deformation moduli of rock mass in
A-19 APPENDIX
 three mutually perpendicular directions are required.
In addition, shear moduli in these directions are also
required. The RQD method provides deformation modulus
only in the vertical direction. Can this method be
extended to determine modulus in horizontal direction
and shear modulus in three mutually perpendicular
directions?
4C 8: The correlations between a modulus reduction factor and
RQD established by Dr. Deere is based on limited field
data and it has many shortcomings; i.e., it is not
applicable for RQD less than 57 and it is not realistic
to use this correlation for rock mass where
discontinuity characteristics are significantly
different from those in the Deere's correlation. Is it
possible to make the correlations realistic by
incorporating field test data gathered from sources
with different discontinuity characteristics?

4D. TUNNELS
Ouestion
4D 1: Has sufficient evidence emerged to recommend RQD
applications or interpretations for very deep problems,
e.g., for rock openings at depths of the order 1,000
meters (as opposed to 10 or 100 meters)?
4D 2: Some engineering geologists have expressed the idea
that RQD, when applied to tunneling, should be
determined from horizontal holes. Is this correct?
4D 3: We have used the RQD index two or three times in past
years to determine an appropriate tunnel support
system. The index was required for analyzing rock mass
behavior with both the Rock Structure Rating Concept
and Bieniawski's Geomechanics Classification. These
qualitative studies have proved to be a very useful way
to describe rock mass quality in addition to practical
experience.

4E. EROSION RESISTANCE, ROUGHNESS COEFFICIENT

Question
4E 1: A question that seems to come up often on both open
channel and tunnel excavations in rock is how to assess
the rock in such excavations for its durability against
"erosion under various velocities of streamflow. The
question we propose is: "Can RQD be used as an
indicator of the durability of rock against streamflow
erosion?" Additionally: "Can RQD be used to predict
the maximum stream velocity a rock mass can withstand?"

A-2 0 APPENDIX
 how to
4E 2: Another problem that seems to come up is on Manning's
estimate the roughness coefficient used in
open
equation for determination of discharge in an bedrock
channel. The problem with an open channel in
due to
is on how to predict what this value might be amount
all the variables such as degree of weathering,RQD be
and orientation of discontinuities, etc. Can
used in some way to help estimate what the value of the
roughness coefficient may be?

A-21 APPENDIX
 5. GENERAL USEFULNESS OF ROD
5A. FAVORABLE EXPERIENCE
Question
5A 1: No problems have been experienced with the application
of RQD as an engineering index. In fact, in the use
for which it was intended, we have found the index to
be a practical parameter for estimating rock core
quality.
5A 2: These remarks are not intended to belittle the
usefulness of the RQD system, but to point out its
shortcomings in practical work. The system has allowed
coordination of the nature of the rock mass to
engineering characteristics in a quick and simple
manner. Any modifications that detract, very much,
from its simplicity would be a disservice. After all,
the natural occurring features of a rock mass are quite
simple and easy to note and provide the basic data
needed for just about any analysis. When experience in
engineering characteristics is added, you have all that
is necessary. What the RQD system does is add the
experience factor for the inexperienced people. Of
course, it does it well, because it draws on a broad
experience base.
5A 3: Like many other index properties, RQD appears to be one
tool available for the evaluation of the behavior of
rock in various engineering applications. This new
study should set RQD in its proper perspective,
including where and how it should be used, and where
and how it should not be used.
5B. SHORTCOMINGS, LIMITATIONS
Question
5B 1: ... We also found during our survey that some of the
largest companies did not use RQD unless specifically
requested by their clients. This was due largely to
the problems with the method as derived in the field
and also with its use in design. Many engineers
apparently will design based strictly on the RQD number
without regard to other factors.
5B 2: ... Most of my experience with the use of RQD has been
favorable but I have noticed through the years an
increase in the misuse and misinterpretation of the
system by engineers not trained in the geotechnical
field. As most of us know, the RQD method is one of
many tools which must be used with other factors to
determine the suitability of the total rock mass.
A-2 2 APPENDIX
 There is a growing number of engineers and architectsRQD
(structural and highway) that have locked on to the into
table without regard to the many factors that went must
the system or the many geological conditions that or
be considered when designing foundations, tunnels,
excavation slopes.
An increasing number of firms are using inexperienced
core drill inspectors who are not trained in good
descriptive logging techniques which result in poor I
rock descriptions and total dependence on RQD values. to
suspect the AE regards the RQD method as a panacea
the rising cost of detailed geotechnical investigations
and rock testing.

RQD is a tool that has been widely used. It can be

5B 3: aids have been
misused by the fact that a lot of design
information on
developed using RQD along with other can
rock properties. With these aids some designers
A discussion on the
take RQD alone to use in design. a
intended use and limitations should be included if
report is prepared.

5B 4: We have had much discussion recently about the

definition and use of RQD, and we would like to As relay
three areas: (1) we
our feelings on the following
understand it, RQD is calculated counting only those
pieces of drill core greater than 4 inches in length.
since
This does not seem to be the appropriate method
it is not a true reflection of the condition of the
of
rock mass in place; (2) the use of RQD by the Corps very
Engineers is confined only to NX core. This is
restrictive since a major portion of the drill holes in
diameter (HQ) than NX; and (3)
the Northwest are larger many
more that it should be. In
RQD seems to be used of
situations, we feel that a graphical representation
is more useful. RQD seems best
the fracture frequency as the
suited for use in rock mass rating schemes such
Geomechanics Classification System, or the Qproperties system,
which is used in estimating the engineering
of the rock mass.
RQD
5B 5: In regard to your inquiry about our use of
measurements, we use it routinely in core drilling.
our
However, I do not think it is that useful in
practice. Generally, I am trying to estimate the
average depth of weathering into rock for foundation
design and cost estimating. That usually involves
pick the
-looking at each core carefully and trying to stops. RQD
depth where any significant rock weathering
does not help much in that regard.
account for
5B 6: I am not sure a simple index property can
these problems. Obviously a study of geologic
A-2 3 APPENDIX
 conditions and core inspection should always be made by
a geotechnical designer, and he should never be tempted
to circumvent this process by relying on simple
parameters prepared by drillers and/or technicians.
Therefore any improvements in RQD must be viewed with
caution if they take emphasis away from performing a
comprehensive evaluation of geologic details that could
adversely effect project performance.
5B 7: A rock quality designation as described by Deere (1968)
provides a means of communicating certain physical
characteristics about bedrock cores. The procedure has
been widely used by geologists and engineers to assess,
in very general terms, the competency of bedrock as it
relates to engineering work. Unfortunately, the
simplicity of the procedure together with the common
practice of correlating numerical RQD with terms like
"poor", "good" and "excellent" often lead to the misuse
and misunderstanding of RQD.
Anyone who uses RQD as a tool of understanding the
engineering properties of bedrock must also understand
the limitations inherent in such a simplistic approach
to assessing the engineering properties of bedrock.
When considered by itself, outside the context of the
local geology, i.e., the lithology of the rock, the
geologic structure of the bedrock and the potential
influence of bedrock weathering, RQD becomes a
meaningless number. Deere (1968) certainly recognizes
the limitations of the RQD system and the necessity for
considering the overall geology of a site when
designing engineering projects that interact with the
bedrock. Therefore, to be properly used, RQD must be
considered as only one small part of the overall
geologic evaluation and cannot be used as the sole
basis for determining the engineering qualities of
bedrock.
5B 8: In summary, RQD serves a purpose and can be a useful
tool to describe certain properties of rock core.
However, its design application is severely limited.
The fact that RQD is correlated with such terms as
"excellent," "good" and "poor," and is sometimes even
-correlated with allowable bearing capacities (Peck,
Hanson, Thornburn, 1974), affords much opportunity for
its misuse. Designers must not rely on RQD alone as a
basis for foundation design decisions. At best, it can
only serve as a tool of limited use in the assessment
of the engineering qualities of bedrock.
5B 9: Problems have been inherent with the system since its
inception and prevent our wholehearted adoption.
First, the elimination of short core pieces from
determining rock quality causes some strong rock

A-24 APPENDIX
 masses, such as, basalts and metamorphics to receive
lower ratings than they deserve. Good examples are the
local diced basalt units which stand in vertical cliffs
(even overhang) due to the irregular nature of the
fracture planes. The RQD system does not incorporate
the attitude of fractures or the presence of clay
fillings along fractures. It does not provide for core
diameters larger or smaller than NX and does not give
an indication of highly fractured zones within core
runs. Unfortunately, some geologists have logged only
RQD values and thought that they had done a meaningful
job of fracture logging. We find that the most
meaningful core fracture logging method notes
individual fractures with their attitude, fillings, and
smoothness. Where fractures are too closely spaced to
be treated individually, as they usually are, then
marking boring logs with brackets or zones and
describing the fractures within the brackets pins down
the weak horizons where they actually are and works out
the best.

5B 10: If the sample condition is not observed upon removal

from the core barrel, during handling and boxing and
immediately logged by a qualified person, the data
presented may be far from the original characteristics.
We often have core samples which are not logged until
after transportation and sometimes days or weeks after
obtained.

Some of our staff feel that, as presently determined

and used, the RQD is nothing more than a statistical
exercise with little useful application in the real
world. Deere could probably spend another 20 years in
developing correction/adjustment values for the
variables which are not actually related to rock
quality but which can greatly affect the RQD numbers.

A-25 APPENDIX
OMLIBMY. LASVEGAS

//Ill)6IIII/IIIIItIIIIIII
010000149968lt