Quantification of Shape,

Angularity, and Surface Texture

of Base Course Materials
Vincent C. Janoo January 1998

:ON STATEMENT K
Kovea tea ptusiic release

19980128 092 [DUG QUALITY mepsioaa) %

I
Abstract: A state-of-the-art review was conducted to de- face texture of coarse aggregates in terms of changes
termine existing test methods for characterizing the in the voids in the aggregate bulk. A description of both
shape, angularity, and surface texture of coarse aggre- the direct and indirect methods is provided in the
gates. The review found direct methods used by geolo- report. Also, the effect of shape, angularity, and surface
gists to determine these characteristics. These methods texture of coarse aggregates on the base course perfor-
involve physical measurements of individual aggregates mance was reviewed. It was found that there is some
and are very laborious and time consuming. Engineers contradiction in the published data on resilient modu-
have developed index tests (indirect methods) to quan- lus. Shape, angularity, and surface texture of coarse
tify the combined effect of the shape, angularity, and sur- aggregates clearly influence the angle of internal friction.

How to get copies of CRREL technical publications:

Department of Defense personnel and contractors may order reports through the Defense Technical Information Center:
DTIC-BR SUITE 0944
8725 JOHN J KINGMAN RD
FT BELVOIRVA 22060-6218
Telephone 1 800 225 3842
E-mail [email protected]
[email protected]
WWW http://www.dtic.dla.mil/
All others may order reports through the National Technical Information Service:
NTIS
5285 PORT ROYAL RD
SPRINGFIELD VA 22161
Telephone 1 703 487 4650
1 703 487 4639 (TDD for the hearing-impaired)
E-mail [email protected]
WWW http://www.fedworld.gov/ntis/ntishome.html
A complete list of all CRREL technical publications is available from:
USACRREL (CECRL-LP)
72 LYME RD
HANOVER NH 03755-1290
Telephone 1 603 646 4338
E-mail [email protected]
For information on all aspects of the Cold Regions Research and Engineering Laboratory, visit our World Wide Web site:
http://www.crrel.usace.army.mil
Special Report 98-1
US Army Corps
of Engineers®
Cold Regions Research &
Engineering Laboratory

Quantification of Shape,
Angularity, and Surface Texture
of Base Course Materials
Vincent C. Janoo January 1998

Prepared for
STATE OF VERMONT AGENCY OF TRANSPORTATION

Approved for public release; distribution is unlimited.

PREFACE
This report was prepared by Dr. Vincent C. Janoo, Research Civil Engineer, Civil Engi-
neering Research Division, Research and Engineering Directorate, U.S. Army Cold Regions
Research and Engineering Laboratory. Funding was provided by the State of Vermont
Agency of Transportation.
Technical review of this report was provided by Christopher Benda, P.E., State of Ver-
mont Agency of Transportation, and by Andrew Dawson, University of Nottingham,
United Kingdom. The author especially thanks Lynette Barna, Sherri Orchino, and Tim
Baldwin for assistance in finding the reports and papers needed for this review, and also
thanks Gioia Cattabriga for editing the report.
The contents of this report are not to be used for advertising or promotional purposes.
Citation of brand names does not constitute an official endorsement or approval of the use
of such commercial products.
CONTENTS
Preface ü
Introduction 1
Characterization of coarse aggregate shape, angularity, and surface roughness 1
Petrological methods 1
Determination of aggregate characteristics using image analysis 5
Determination of aggregate characteristics from indirect methods 9
Laboratory response of angular material 16
Summary and conclusions 20
Recommendations 20
Literature cited 21
Abstract 23

ILLUSTRATIONS
Figure
1. Aggregate classification chart 2
2. Roundness chart for 16- to 32-mm aggregates 3
3. Parameters for determining the degree of angularity 3
4. Degree of angularity chart 4
5. Visual identification of aggregate angularity 4
6. Measurement method for characterizing the surface texture of an aggregate 4
7. Image of the Pike crushed stone prior to image analysis 5
8. Image of the crushed stone prior to image analysis 5
9. Image of the crushed gravel prior to image analysis 5
10. Material characterization using image analysis 6
11. Definition of roughness and roundness 6
12. Relationship between mean aggregate diameter and porosity 10
13. Packing volume, packing porosity, and geometric irregularity of aggregate
particles 13
14. Schematic description of the pouring device 14
15. Packing specific gravity and rugosity for different aggregates 14
16. Test apparatus for ASTM C 1252 and modified ASTM C 1252 15
17. QMOT time index test apparatus 15
18. Effect of crushed particles on cyclic creep strain 17
19. Effect of combined sphericity and angularity on cyclic creep of base course
materials 17
20. Effect of sphericity and angularity on the angle of internal friction 18
21. Shape classification of aggregates 18
22. Influence of material type and state of stress on resilient modulus 19
23. Influence of material type, stress level, and percent fines on plastic strain 20

iii
TABLES
Table
1. Description of aggregate shape 2
2. Criteria for describing particle shape 2
3. Criteria for describing angularity of coarse-grained particles 4
4. Image analysis results for crushed gravel 7
5. Image analysis results for crushed stone 8
6. Image analysis results for Pike crushed stone 8
7. Mean and variance of roughness and roundness values from image analysis 9
8. Value of shape factor for various materials 10
9. Results of particle index tests for various aggregates 11
10. Mold size and dimensions of tamping rods for the particle index test 12
11. Dimensions of mold and tamping rods 12
12. Specific setups for coarse fractions for the pouring test 14
13. Average flow coefficients for Vermont subbase materials 15
14. Summary of factors affecting <j) 16
15. Gradation of test aggregates 17
16. AIF values of test aggregates 19

IV
 Quantification of Shape, Angularity, and
Surface Texture of Base Course Materials
VINCENT C. JANOO

INTRODUCTION irregularities and the effect of these irregularities

The base course layer is identified as that un- on the performance of base course materials. A
bound layer between the surface course (asphalt summary of the methods used by geologists to
concrete or portland cement concrete) and the describe aggregates has been included. Even
subgrade. This layer is sometimes subdivided though these methods may not be practical for
into two parts: the upper layer, called the base, everyday engineering work, they have been used
and the bottom, the subbase. The base is usually by researchers for quantifying base course perfor-
constructed with coarse gravel or crushed stone, mance (Barksdale and Itani 1994). A review of
and the subbase with either gravel or sand. Sever- current methods used by engineers, including a
al factors affect the performance of these layers. discussion on the usage of image analysis for de-
The common factors usually cited are gradation, scribing particle shape and roughness, is also pre-
degree of compaction, amount of fines, moisture sented.
content, and permeability. The shape, angularity,
and surface texture of the aggregates also have a CHARACTERIZATION OF COARSE
significant effect on the engineering response of AGGREGATE SHAPE, ANGULARITY,
the layer. AND SURFACE ROUGHNESS
Quantification of the shape, angularity, and
surface texture of aggregates is difficult but not The description of the aggregate particles can
impossible. There are several methods that in- be conducted directly using petrological methods
volve either direct measurement of the aggre- or by using image analysis. The effect of particle
gates or indirect inference from aggregate proper- shape, angularity, and surface texture on the
ties. For example, geologists have developed a engineering properties can also be inferred.
sophisticated system that involves physical mea-
surements of the aggregates, and engineers have Petrological methods
developed visual classification methods for char- The shape of the aggregate can be described by
acterization, shape, and angularity or index tests its length, width, thickness, sphericity, round-
that use some engineering properties, such as po- ness, and angularity. The surface texture is diffi-
rosity, to speculate on the shape, angularity, and cult to quantify and several visual charts and
surface texture of aggregates. However, none of some sophisticated test methods have been pro-
the index tests developed by engineers make it posed. A description of these methods is presented
possible to quantify separately the shape, angu- below.
larity, or surface texture. Usually these character-
istics are lumped together as geometric irregular- Description of aggregate shape
ities. Sedimentary petrologists have described the
This report reviews test methods currently shape of particles using methods developed as
available to quantify the aggregate geometric long ago as 1930. The description involves mea-
surement of the aggregate dimensions (shortest, elongated using the criteria shown in Table 2.
intermediate, and longest lengths). Based on These criteria were developed basically for iden-
these lengths, the shape of the aggregate is quan- tifying flat and elongated aggregate and not as a
tified in terms of both the flatness and elongation general tool for describing the shape of coarse ag-
ratios or by the shape factor. The flatness ratio (p) gregates.
is the ratio of the short length to intermediate
length, and the elongation ratio (q) is the ratio of Description of roundness
the intermediate length to the longest length. Roundness, as opposed to shape, has to do
An alternate method for describing the shape with whether the edges are sharp or round
of the aggregate involves the shape factor (F) and (Wadell 1932). The roundness of a particle is sen-
sphericity (\|/) of the aggregate particles. The sitive to abrasion and wear to which aggregates
shape factor (F) is the ratio of the elongation ratio are subjected during processing and construc-
(q) and the flatness ratio (p), i.e., tion. The process for determining the roundness
of a particle is fairly complex and involves tracing
a magnified image of the particle. An inscribed
circle is drawn within the aggregate tracing and
the radius of the circle is determined. The radii of
A shape factor equal to 1 represents a rounded
or cubic aggregate. Smaller than 1, the particle
tends towards being elongated and thin. Table 1. Description of aggregate shape.
Blade-shaped materials have a shape factor
Aggregate shape Description
greater than 1.
The sphericity (V|/) is defined as the ratio of Disc Slabby in appearance, but not elongated
the surface area of a sphere having the same Equidimensional Neither slabby appearance nor elongated
Blade Slabby appearance
volume as the particle in question to the sur- Rod Elongated, but not slabby in appearance
face area of the particle. The sphericity can also
be determined from the flatness and elonga-
tion ratios as shown below: Table 2. Criteria for describing particle shape.
Shape Description

y. Flat Particles with width/ thickness > 3

2 2 Elongated Particles with length/width > 3
l + p(l + q) + 6jl + p {l + q ) Flat and elongated Particles that meet criteria for both
flat and elongated
The shape factor, p, q, and sphericity are used
to classify the aggregate as disc, equidimen-
sional, blade- or rod-shaped using Figure 1.
Equidimensional can mean either cubic or
round. The definitions for disc, equidimen-
sional, blade, and rod, based on descriptions
by Lees (1964) and Barksdale and Itani (1994),
are presented in Table 1. II
O)
Barksdale and Itani (1994), in their study on C
O
_l
the effect of shape, angularity, and roughness
of aggregates on base course performance,
used this form of shape description for charac- (D
E
terizing the five aggregates they used. The B
granite gneiss, limestone, shale, quartzite, and
river gravel turned out to be disc, blade, blade,
blade, and equidimensional in shape, respec-
tively. Additional discussion on the results of
his study is presented later in this report.
ASTM D 2488-90 (1996), Standard Practice for
0.2 0.4 0.6 0.8
Description and Identification of Soils (Visual-
Short/Intermediate = p
Manual Procedure), describes the shape of ag-
gregates as either flat or elongated, or flat and Figure 1. Aggregate classification chart.
 Wf
RWfti*SS»~ftJ: Round»«» = OJS nxA«s Sij
001 = «1.4 f*0W*WS = U

«US ft*

494»
ft» a.4

ftMKteasa s'ftB Fta««a»ass=&S Aoundnossi s OLD

Figure 2. Roundness chart for 16- to 32-mm aggregates. (After Krumbein 1941.)

the edges and corners of the particles also are crushed aggregates. Lees (1964) proposed a meth-
determined. These radii are used to index the od for determining the degree of angularity, which
roundness of the aggregate as shown below. accounts not only for the roundness of the corner
Details on this method can be found in the papers but also how far the projection is from the in-
by Wadell (1932) and Krumbein (1941). Essen- scribed circle (Fig. 3). The degree of angularity is
tially, the roundness (R) of a particle is an average calculated by the following equation:
measure of curvature of the corners and edges. R
is expressed as A{ = (180° ■«)-

r
where A\ = degree of angularity
R = —r- a= measured angle
N x= distance to the tip of the corner from
where r = radius of curvature of a corner of the the center of the maximum inscribed
particle surface circle
r' = radius of the maximum inscribed cir- r= radius of the maximum inscribed circle.
cle in the projected plane The total degree of angularity (A) is the sum of all
N = number of corners. the values for all corners measured in three mutu-
Because this is a lengthy process, visual charts ally perpendicular planes. Again, because of the
have been developed (Krumbein 1941) for esti- high degree of complexity, Lees (1964) developed a
mating the roundness of aggregates. Figure 2 can visual chart for determining the degree of angular-
be used to classify the roundness for 16- to 32-mm ity of particles (Fig. 4).
aggregates. By enlarging or reducing the chart, ASTM D 2488-90 (1996), Standard Practice for De-
similar roundness charts could be made for other scription and Identification of Soils (Visual-Manual
size aggregates (Krumbein 1941). The roundness Procedure), describes the angularity of coarse-
(R) varies between 0.1 and 0.9. An R value greater grained material as either angular, subangular,
than 0.6 indicates high roundness, R between 0.4
and 0.6 indicates medium roundness, and R less
than 0.4, low roundness.

Description of angularity
Lees (1964) found that when using Krumbein's
(1941) method for determining the roundness
number, it was possible to get the same round-
ness number for two very differently shaped Figure 3. Parameters for determining the degree of
aggregates. He found to it to be especially true for angularity. (After Lees 1964.)
 0-99 100-199 200-299 300-399 400-499 500-599 600-699 700-799

oo oo oo
oo ÖD
00 oO Q0 Oo
oo 00 OO o a OO
Do
800-899 900-999 1000-1099 1100-1199 1200-1299 1300-1399 1400-1499 1500-1599

^7 c a Ö
8°
Figure 4. Degree of angularity chart. (After Lees 1964.)

Table 3. Criteria for describing angularity of

coarse-grained particles (ASTM D 2488-90).
Description Criteria
Angular Particles have sharp edges and relatively ■Jjl ftasctaiJ
plane sides with unpolished surfaces.
Subangular Particles are similar to angular description
but have rounded edges.
Subrounded Particles have nearly plane sides but have
well-rounded corners and edges.
Rounded Particles have smoothly curved sides and
no edges.
Figure 5. Visual identification of aggregate angularity.
(After ASTM D 2488-90.)
subrounded, or rounded. The qualitative criteria,
presented in Table 3, are used for describing the
angularity. A visual chart for determining the scale to determine the surface texture (roughness)
angularity of the aggregate, similar to that found of aggregates. The surface roughness (SR) was
in ASTM D 2488-90, is presented in Figure 5. based on visual inspection and indexed to a scale
ranging from 0 (for glassy particles) to 1000 (for
Description of surface texture very rough particles).
The surface characteristics or texture of an Another method found in the literature was to
aggregate are considered to have an effect on the coat a flat, sawed aggregate surface with asphalt
engineering response of the material. Terzaghi (Bikerman 1964), which is then scraped to the sur-
and Peck (1967) defined texture as the degree of
fineness and uniformity. They suggested qualita-
tive expressions such as smooth, sharp, gritty,
etc., to describe the texture.
Barksdale and Itani (1994) used a roughness

8 cm chords
4 cm chords

Figure 6. Measurement method for characterizing the surface texture of

an aggregate. (After Wright 1955.)
face of the aggregate. The remaining asphalt on the Vermont Agency of Transportation (VAOT)
the aggregate is used as an indicator of the sur- were characterized using the QMOT image analy-
face texture. This method determines the micro- sis system. The samples included one sample of
texture of the surface and the absorptive nature of crushed gravel and two samples of crushed
the aggregate and is still an index to the rough- stone. The crushed stone was obtained from two
ness of the surface. different sources. Roughly thirty to forty ran-
Wright (1955) developed a method for quanti- domly selected aggregates were placed on a light
fying the surface texture of concrete aggregate table and videotaped. The outlines of aggregates
particles using studies done on 19-mm stones. from the three sources, as reproduced by the im-
The test aggregates were first embedded in a syn- age analysis program, are shown in Figures 7 to 9.
thetic resin. Once the resin hardened, the stones
were cut into thin sections (approximately 2.5 x
10~2 mm thick). The thin section was placed un-
der a projection microscope and magnified 125
times. The unevenness of the surface was traced
and the total length of the trace was measured.
This length was then compared with an uneven
line drawn as a series of chords (Fig. 6). The
difference between these two lines was defined as
the roughness factor. The reproducibility of this
method was reported to be good (Wright 1955).
However, to obtain a reliable average of rough-
ness of any one aggregate size, it is necessary to
take many sample measurements. Figure 7. Image of the Pike crushed stone
prior to image analysis.
Determination of aggregate characteristics
using image analysis
The direct measurement methods described in
the previous sections are time consuming and
labor intensive. Thanks to the power of present
computing capabilities, computer-based image
analysis has been developed for analyzing aggre-
gate shape, angularity, and roughness. A feasibil-
ity study was conducted for the Federal Highway
Administration (FHWA) on this technology (Wil-
son et al. 1995). The study was conducted on fine
sands (manufactured versus natural), and the
conclusion was that image analysis was a viable
tool for distinguishing the shape and angularity Figure 8. Image of the crushed stone prior
to image analysis.
of the particles. Wilson et al. (1995) developed the
Roundness and Shape Indices to quantify the
roundness and angularity of the particles, re-
spectively. The method first involves capturing
images of the aggregates using a high-resolution
video camera. Then an image analysis computer
program is used to identify, separate, and trace
the edges of the aggregates. Based on the traces,
algorithms in the programs are used to determine
the different characteristics of the aggregates.
The Quebec Ministry of Transportation
(QMOT) uses image analysis on a routine basis
for verifying the angularity of its hot mix aggre-
gates. To study the feasibility of this method, Figure 9. Image of the crushed gravel prior
samples of base course aggregates obtained from to image analysis.
 s—^

^Hv
\ -■—"-—'

±y- r
^--N ^—^ ^r
1\ T \
V. V
~>. _, „ -<.. V
:
^S'
• ^—*
"'iy^ ^"v
^^
X
/

a. Area o/an aggregate calculated through image analysis. c. Typical feret measurement.
l^^
N
/ \ s**"^
t x y s
X ™ ~r>-t
\ ^
^ V-
4 ^ t
Z J v. ^
i~t ■ ^^" '•' ■ \_
' "-t,J;
b. Perimeter ofan aggregate calculated through image anal- d. Illustration of convex perimeter.
ysis.
Figure 10. Material characterization using image analysis.

area of an aggregate is the sum of pixels present

within the aggregate boundary (Fig. 10a). The
perimeter is the sum of pixels surrounding the
aggregate (Fig. 10b). Ferets are straight-line meas-
urements made between two tangents (Fig. 10c).
In this particular case, eight feret measurements
are taken at 0,22.5,45,67.5,90,112.5,135, and 157.5
degrees. From these measurements, indices for
roundness and roughness are derived. For exam-
ple, the length and width of the aggregate are the
longest and shortest feret measurements taken at
the various angles mentioned above, respectively.
The roughness is defined as the ratio of the
perimeter to the convex perimeter. The convex
perimeter is defined as the string measurement
b. Roundness. around the tips of the eight ferets (Fig. lOd). For a
Figure 11. Definition of roughness and roundness. smooth material, the roughness factor equals
1.00. As the roughness increases, the roughness
factor also increases (Fig. 11a).
The numbers on these figures are used to iden-
The roundness is defined as follows:
tify the measurements of the individual aggre-
gates during the analysis process. The following 47L4
basic measurements were made: 1) feature area, Kn =
2) feature perimeter, 3) feature ferets, and 4) x-y
coordinates. In the image analysis method, the where Rn = roundness index
 A = area of aggregate results appear to be close to one another and it is
P = perimeter of aggregate. unclear at this time whether this is the range one
can expect between crushed gravel and crushed
The roundness index for a perfect circle is 100. stone. Visual examination of the crushed gravel
However, with image analysis, because of the and crushed stone showed that the crushed
square pixels matrix, the best that can be obtained gravel mostly had one fractured face with the re-
is 95. As the material becomes angular, the round- mainder of the body being rounded. The crushed
ness index decreases (Fig. lib). stone had several fractured faces and was defi-
The results of the tests conducted on the VAOT nitely more angular than the crushed gravel.
aggregates are shown in Tables 4 to 6. The mean Statistically, however, at the 95% confidence
and variance of the roundness and roughness for level, the roundness mean value between
the three aggregates are shown in Table 7. On the crushed gravel and crushed stone is significantly
average, the results from crushed gravel as com- different. The mean roundness values between
pared to the results from crushed stone indicate a the crushed stones are not significantly different
decrease in roundness and an increase in rough- at the 95% confidence level.
ness, suggesting that the crushed stone is more The mean roughness values ranged from 1.031
angular and rougher. The roundness index to 1.050 for the crushed gravel and crushed stone,
ranged from 78.54 to 72.36 for crushed gravel to respectively. Again, the range is fairly small, even
crushed stone, respectively, indicating by defini- though the crushed gravel was fairly smooth and
tion that all the materials are fairly angular. The the crushed stone was rough. Statistically there is

Table 4. Image analysis results for crushed gravel.

Calibration: 0.2639 mm/pixel
Aggregate Length Width Breadth Area Perimeter Roundness Roughness

1 19.6069 11.9607 13.9666 184.3518 55.0471 76.4519 1.0379

2 17.3974 13.5917 15.1222 171.3743 51.4075 81.4897 1.0371
3 32.8975 22.6729 24.7369 523.6398 90.942 79.5635 1.0365
4 8.8634 7.0677 8.0576 45.2048 25.7946 85.3764 1.0106
5 20.6594 18.0406 18.8365 284.5665 63.7988 87.8554 1.0327
6 24.9787 21.487 21.487 378.1483 76.026 82.2143 1.0298
7 29.8133 19.0284 23.0986 446.1356 84.5824 78.364 1.0295
8 19.8755 9.4006 9.4006 143.617 51.2797 68.6316 1.0258
9 19.2669 12.8389 12.8389 194.0849 53.7019 84.5712 1.0193
10 20.1157 14.7801 15.1222 204.8994 56.8332 79.7161 1.0421
11 11.6129 8.155 9.834 66.8338 35.615 66.2125 1.0762
12 15.4945 9.2375 10.0815 90.6258 40.2236 70.3881 1.0132
13 20.4127 17.1896 17.1896 256.1603 61.3262 85.5913 1.0288
14 30.8876 20.9589 22.83 436.9793 84.7282 76.4918 1.0432
15 27.4552 20.4127 21.2242 380.6716 80.5014 73.8164 1.0377
16 8.8634 7.7891 7.7891 47.0072 25.8876 88.1435 1.0012
17 17.4827 14.3263 15.8358 174.2582 52.825 78.4736 1.0339
18 28.7389 24.2815 27.1274 520.9001 89.8696 81.0474 1.0431
19 17.7753 7.6114 7.6114 103.7474 46.2675 60.9026 1.0454
20 29.2761 23.6358 23.6358 380.3112 85.3504 65.6051 1.0458
21 21.7556 18.5712 20.6813 296.1741 66.4352 84.3258 1.0346
22 13.9666 12.6686 13.4294 121.5554 43.8107 79.5837 1.038
23 18.4751 11.1999 11.1999 159.5504 50.0157 80.1486 1.0175
24 12.1408 7.3755 7.3755 69.6456 32.8435 81.1344 1.0132
25 28.7389 16.6525 16.6525 342.4603 76.8337 72.8982 1.042
26 23.0986 15.0409 15.0409 264.5235 66.0363 76.2269 1.0287
27 10.8734 7.1632 7.1632 50.4678 29.1575 74.5975 1.0173
28 18.2129 15.8467 17.2447 220.7607 56.0032 88.4516 1.0181
29 12.6236 9.8162 10.4749 85.3627 36.7793 79.2995 1.0404
30 21.0339 16.1835 17.9472 242.8945 61.7914 79.9413 1.0414
31 28.6527 17.1255 17.1255 352.1213 77.1649 74.3126 1.0277
32 12.8389 7.6938 8.9736 71.5922 34.6879 74.7685 1.0324
33 9.4006 7.6938 8.0576 48.5212 26.73 85.3384 1.0048
34 9.5015 8.155 8.4682 57.3891 28.5749 88.3224 1.0175
 in in co vo H n CM CO t-t CO ON M ^ ^ rj 0\ N CO ON 00 CO CM "<tf CM VO COCO^O^CO^OsONOONtTlN
O rH CM CM rH CM O cNincoirio>^csi/)iococriN"*csiN(r)'fHNHfsiOHN\oNHOcoHiriHo^o\rt
^ in \o m 51 ^ -^ r^r^rtr^r^r^c5C!ic!i<^(^c^c^C!5C5CDCDC3C3000c^oC)C>CDOOOOOOOOO
o o o o o o o OOOOOOOOO

N tN co ^ ON *# in MifiON^innrnco^N vo IN m 00 00 (N IN IN IN M< 00 rH
oi cOrHcMobsOocoo Fs' ^j S (S is -----
co iri
KvoincooNinoooinvprH
vo co ^N^^vOlfltON
_._,__ _ .... _
*3<
M H n (n ON w in 5 CM
•tf vn o IN rH
ro m rH 00
a\ rH VO m IN
a> in m in vr>
o
3o 00
00 rf 00 IN
fc
a inovoinr-icN'-otN.a-vO tNcocoin^covooNcocMcococotN IN m vo O m CM tN ■* Ov «3 rH VO O in r~^
o
in
CM IN K K IN
CO vO v£> VO VO
00 00
in IN a CJ\ 00 O
VO IN VO R fc s fc 00
VO EN KR 00

V
in vo o* in vo vo oo co ON ov H ^ OVHHIHWIvl^HOrtrJI tNooinvo^vocoiNovin nMiNNmioN«
-*oiHinxrt^Ham w. ... tNtNOOOoorHoinovinoo OCNrHOOvVOCMlNmOOCTvOOrHCOiniNrHCNav
tNtNOb —
in-tfotNcovocritN^foavoavvo „,Ä OINmtNVOOVOinavOOOincMOVCMCOVOOVO
QOHNON^io^ovo\O)incocqNiON0jcnNin^qcoNaoHisinNcoov\Oino)O;voq^tno^
iN#voco^o>o>o>LncqcocMincNa\cncNiinjÄ" cDcMovorHcNincvicooNvomavovvoo'^mo^
"
rHrHcoo>cviininrHÖrt\öiNÖo>övd^rHcöcö^o>inöo6KovtNincM
rHCOCVCNlrilrirHOPSvpiN
01
&
OV CM in K VO co in -rt^ONH'f'vtOAN^^COCOH
VO ^ ON *<* m rH T}*
n CM CM VO tN OOCOCMtnrHCMCOOvOOCMrH K CO
^S* *st^ CO l"N1
m «*
I'J *tf
VJ' CM I >| L »» \J> ^
V>| CM -*J' 00
w CO -

CM .-, ~. ^ K.,
CO CM vO rH VOin-sfvOCOrHCMCM-fvOCO oocoLnvotNvoinvototNONON'^
n ,.., ~, ^ *., ^, ~, „ .... * , w. -. , (NCOinCONvONvOmHCOCnONCO
o Ö tvnqn CM TfNO vq»iON\ON\ONinin^qrjO)qoo\OHONCvj\oijq^Nqco
*,w
vö ö co a\ IN örHCOCMCOtNCOrHÖKoNCÖÖlrilN^^rHlricOvÖtN^
IN Ov O CM incOOCO^(SCOONNlf)0>NNHvONO\tr)\00\HQOCOlOHHVOCONOH\OlNONHO>^ts
V) n CMCNrHrHrHCMmCMrHrH rH rH rH *~H CM rH rH rH rH rH rH rH rH rH CM
o
3
a .O.

u
a
VO CO "* rH ^ HO in mmi-irscocNpNtxco
01 COCMCOCMONCOVOOOCO CMONrHONCMCSCOONONCNlCOCOCOrH -vfCNONCOO^OvlTl, rf CM tN rH CO CO CO
6D
rt
coococooorHcoo NN*OHHONOOvDHN(NnC)CMri'T(NcOOHOMCONOrO^OO^(N
----- ■■ — - — — -i ■- -• .- — —
in^vON^CONNH CMrHONO>COOvrH^COvqpCO(NincMU^ONr^ HcNcoHinN^MiovorjN^incoa
fÖcÖONOvMiONlfi^H cNinoKvdcMiricöcooNcocNioNrH^cocöcMO inrHCOlNCOrHcKOCOrHCOcOONt^t^
E
vd
01 CO CO if) co in N ■^ ON rH
ONv0KrlC0C0NH^^NC0NNC0NOinNCMHN0NNC0'*rf)Nm in -, rH
. . O CM o ^
„ _„ _„ o ■*
in _i_u « ON VO IN
Ä. «^
oocx>vOLnoNONONONrHpcocoincMinrHcoincovoin^co^ ON CO VO VO "' N <* rt< N CO _«
^
\Oin'uriWrHrH|s!\O0ÖT#CTNV0'^CT\vOC^CM^
rHrHrHrHrHrHrHCMrHrHrHCM^rHrHrHrHrHrHrHrHrHrHCMCN]rHCMrHrHrHrHrHrHC^ rH

HNfO^invONMONOHlNcn^invON
CM a1

co in vo K vooNcovom ^oNHHN'tinN CM
3 CO CM ON CM ON, ^O ts.
ThONO>t%inoNOinoo\ococMrHco^inrHONinvoco
CO CO tx ^ CM VO CO
Ei VOCOrHONONCMOONONOKvO-^OVOrH
tovovovoHT)*vOMNvo^^ioin,NHin
ooooooooooeä o — o
— o
— o
— o—

«
CMNOQOONVOCOOO ONCO CO rH VO l>s ON 00 VO cocMrHininvotN.t\oo
OrHONvocMininco
N to H co N J tX VO CM VO T}* in rH
NOANOOHVOVOCOOH..,..OcOvOONCMOONin
_ rHrHONinrHCMvOONOOONONt^COincMCOt
----- CM CO CO ~
cMixmoocooNcn-^cM'^corHCMin ^NH\OCOir)nCOH^vONN00\ON^ . CM^OVOONCMOVO
_. .
TflCNCNNNONCSr- -' -' voint^ininrHcoovooNcocHpONint^vq'^'^ ■■ ■■ IN vo o co in m
01
d
COtNrHrHCÖcÖCOÖÖ0bCMONin*^C0 CM Ö CM
NvOVOCOtstsNNQOVONvOKNN CO N CO N N VO
CO
vo
N
co
N
vo
CO
INEi
CN
t^s
CO N
vo tn
o
•M
<A
oocooNtn^cMooco
cMrHinixoocooNtn'^cMooco COONCO vooN^cocorHcOrHvo
»OOV^COCOHQOHVO Hir)incooiifiiriHHtoo\NNN
r^lGiß
T< VO r-c CO ON
voNonirjinNcorv|in^HKcoH^oo\ONHOHrtcocoNvovoo^N(ONoocoLf)
N^awiovoNo«ifiinr>icorv|invOrtKcoH^oo^NHgHrtcocp
ON r^ ts. O ^ t>*
01 oincMONCMOcMcovovorHoooNCMcovo^OrHininu-iin'N^ovoo^ts.voo
43 cqinr^rHCMoorhcMrNvovopcor^pcMco^cMtxONV^
O^^^ONCOinr^rHCMCO^fNKvOvOpCO^pCMW^ .
tn X icOCOCMCOCNKKcÖKÖÖ^CNKONCMKcOtNrHO^
CÖrHincÖcMCOCOCMCOCNtNKcÖKÖCJ^CNKaN
Ru 0-
iN^ir)N^\Drt^^N^NinrovDNirjcv|ro^N^criinn^^cO'tcoco
voinwNcoN^inrN^vocoV^S^N^ro^

h
O
MH
I
**
ON
VO in rH CM CM ONtnoooNco
ONOOinCM^CN^COincO^OrHgNKrHvOtxONCM vo vo CM N\OHCOOMCOO\0\tvlHCOONvOCOO\CMCOmNCO^O\
^^HM?ioifiroinroMnnoo\NNHco^ VO ON IN cMvo^coinrHinoNinoincjOrH^aNrhoincocorHcooN'N^
vocviojNH'vt'NiritriifiifiH^coNinon^oN ^-«.; VO CO rH
^,
co co rHtx tniP
-, » j u ^ v -. U J >_l 1 U J N_./ u 1 irfV < ■ -J ■ \J 1 Nl' v_rf " J ww •- ■ J »-■_' CO ON VJ|
A^t **. ■ h. ^^ ^JJ ■ *n i^v i^* b-ii n^t r^ ^\ #^l r*r\ rv^s
u rHoocM^o^rHcoinin^oqoqpr^oqcOrHcovoo^K'vf j #^\ v^\ /^. rs. <m-M
H-» IM COvOHvÖcO^ON«vDdcOvÖNvD^d\Ovddl0^tNi
HHin'o>^coövo^voco^cMrt5öuSininööcMvxjin ^in^ONvOvOrHO^OvpONrHCOCMtNint^COpCOtNON
3 CO^in^0^rHinCOONVOKrHCOONOrHl>sCOCOtNOCO
CMrHCOCO^ rHrH CM rH,-(CMrHCOrH CNCNCM rH rH CO
i/)
01 s
O
in co co in in COVO VOCOrHVOONCO CO^rHVO COCOCMCMCOCMCOin in co co
_.....___00 vo CM CM CO
----- -^- "4" ON CT\
._=.=. CO
ts. O tX 00 CO vo CO ^NCONlOrOHvOOCOlONOCONNNvOHNvOCNN^inNÖ^CNlp^ONCONCOON^
N N O N CO CO CO QOinninconNn^rtwcNoonoHHWNwinHNQjacnojHNrt
VO vo CO in rH \0 CM O^ON^COOOOO^OONVOCqO^rHCOCMCM^^VOOCMincOtNOqONC^
IS tNQ>^ONCMCOlrirHK^ONO^OvÖvOONONC»Tt'ÖcÖ
rH rHrHrHrHrHrH i—I CM rH rHi-HrH rHrH rHCMrH rH rHrHrHrH
C
rs
01 tJ u c
00 in T* en vo
vDVDinavrpvü
vo in ON -* vo M to H o\ co to m NONONcocooNCOcocoN ^ ^ O_uco.c^?yc^IC5c5£2c^ON
IS NvoKco^ocooo^^oocM^rHvovooinvooNtNinrHincM
KvOoKKKcOCOinrHCOin^rH^cOVOincOCÄt^CM^^rHVOCOCOrH
VOOCOinvOOOCMCOrlHtNO^^rHOOVO^Ot^vOrHCMC»
in'rHC»OÖ0ÖvÖrH'rHvÖcÖ^a<ÖrH^rHt<^CÖ0Ö
rHrHrHrHrH t-HrH rH rHrHrHrHrHrH rHt-HrH rHrHrH rH H (N| H rH rHrHt-HrH
in
01
'NfONCMincovoiN.ONCMin miOHionchvoHONH \o 10 co jo N 3 00 CO CO CO rH ON CM in in
■^ONCOCN.ONbs.rHCOts.rHOVÜVO vS
oincMininoNinincocörHCMOcMrH^ovcöcocMincocMCM^^co in co
VOOlNWCoioOOH'*ONQON'tOlNtOCO(OON^OOOOOtONNCO in co COHHCN^COCOtOVONvOVO'vt
E2 NO\io^toinoo\\OHiso\^NHcqo\qcotsqco^\qoNN tN co - CM
Ov -■-■-^NiDNvotsvqcqvq
CM VO
HcoH^doN^co^iri^N^vo^co^NcÖH^ONCon^dincoovcv! in CM co in ts^fNNlficÖtNO^
CM H fO N CO rHrH HNHHHNHMHHNCJH rHrHrHrHrHrHrH CM rH rH rH

HNCO^lDvONCOOAOHMCO^lDVON
CM a
Table 7. Mean and variance of roughness and gold characterized the angularity of all other ag-
roundness values from image analysis. gregates as the difference between the percentage
Roughness Roundness of voids and 33%. Then, AN was found to range
Aggregate type Mean Variance Mean Variance between 0 and 12:
Crushed gravel 1.0306 0.000211 78.537 46.668
AN = percentage of voids - 33
Crushed stone 1.0500 0.000198 73.285 37.846
Pike crushed stone 1.0471 0.000115 72.365 66.589
100M
AN = 67--
a difference between the crushed gravel and cG,
crushed stone. There was no difference in the
where M = mass of standard volume of aggre-
mean roughness values between the crushed
gate (g)
stones. From these very limited tests on coarse
c = mass of water required to fill the
aggregates, and without additional strength/
same volume (g)
modulus related testing, it is not certain whether
Ga = oven-dried specific gravity of the
image analysis can clearly distinguish the shape,
aggregate.
angularity, and roughness of different aggre-
gates. Several limitations of the angularity number
were pointed out by Lees (1964). Primarily, the
Determination of aggregate characteristics AN was developed based on results from only six
from indirect methods samples of coarse aggregate. Furthermore, the
It is clear from the above discussions that the angularity as described by Shergold, "angular to
determination of aggregate shape, angularity, rounded," was based on a consensus reached
and surface texture is a fairly lengthy and labori- from visual examination of the test aggregate by
ous task. An alternate approach taken by engi- 25 observers. Another quirk with the AN was that
neers is to infer these characteristics from the it was not to be applicable to all shapes (such as
mass properties of the aggregates. Several indices regular geometric objects [spheres and cubes]),
for coarse aggregates, such as angularity number, and the AN for perfect spheres was found to be
particle index, rugosity, uncompacted void, and higher than for perfect cubes, which appears to be
time index have been identified in the literature. contradictory. Furthermore, the heavy compac-
tion required may break the aggregates, causing
Angularity number artificial changes in angularity.
The angularity number (AN) developed by Gupta (1985) refined the AN model to account
Shergold (1953) is recommended by British Stan- for the shape of the aggregate. The AN was calcu-
dards (BS 8121975) for indexing the angularity of lated in the same manner as before, with the ex-
natural and crushed aggregates used in concrete. ception that Gupta defined the percentage of
Shergold found that when the aggregates were voids in an aggregate mass as a function of the
compacted in a prescribed manner, the percent- shape and the average size of the aggregate
age of voids in the aggregate mass decreased as (mm). Based on test results of the three different
the aggregates became more rounded. He also aggregates, angular limestone (6- to 50-mm
found that as the amount of round gravel in- range), subangular crushed quartzite (6- to 50-
creased in a mixture of natural and crushed ag- mm range), and rounded gravel (6- to 100-mm
gregates, the percentage of voids also decreased. range), as well as different proportions of limestone
Based on his study of six aggregates, he found and rounded gravel mixtures, he found that the
that the minimum percentage of voids in round- percentage of voids (Fig. 12) can be expressed as
ed gravel was approximately 33%. The tests were
conducted on 19-, 12.7-, 9.5-, 6.35-, and 4.76-mm vi = Cdn
aggregates. The test procedure involved compac-
where r| = percentage of voids
tion of individually sized aggregates in three lay-
C = shape factor
ers in a 2800-cm3 mold. Each layer was compact-
d = volume mean aggregate diameter
ed with a tamping rod that weighed between 900
(mm)
and 950 grams to 100 blows. The percentage of
n = exponent.
voids was calculated by using the net weight of
the aggregate in the mold. The size (d) is determined by taking a known
Using round gravel as a reference point, Sher- number of particles and soaking them in water
 rii
o

Log, d
Limestone
Limestone & Rounded Gravel (0.875:0.125 v/v)
Limestone & Rounded Gravel (0.750:0.250 v/v)
Limestone & Rounded Gravel (0.625:0.375 v/v)
Limestone & Rounded Gravel (0.500:0.500 v/v)
Limestone & Rounded Gravel (0.333:0.667 v/v)
Limestone & Rounded Gravel (0.167:0.835 v/v)
Crushed Quartzite
Rounded Gravel

Figure 12. Relationship between mean aggregate diameter and porosity. (After Gupta 1985.)

for 24 hours. The amount of water displaced by Table 8. Value of shape factor (C) for various mate-
these particles is the volume of the aggregates. rials. (After Gupta 1985.)
Gupta (1985) divides this volume by the number Materials Value ofC
of particles to get the mean volume size of the in-
dividual particles. The mean volume size is then Angular limestone 45.43
Subangular crushed quartzite 41.44
converted to the volume mean diameter. The ex- Rounded gravel 37.89
ponent (n) was found to be a constant equal to Limestone + rounded gravel (0.875 :0.125 v/v) 43.49
-0.032 for all the materials and mixtures tested. Limestone + rounded gravel (0.75 : 0.25 v/v) 42.63
The coefficient C was found to be a function of Limestone + rounded gravel (0.50 :0.50 v/v) 41.98
the aggregate shape. It was also found to be a Limestone + rounded gravel (0.333:0.667v/v) 40.58
Limestone + rounded gravel (0.167: 0.733v/v) 39.40
function of the amount of rounded gravel in the
mixture. Values of C as reported by Gupta (1985)
are presented in Table 8. The coefficient C ranged Particle index
between 39 and 45 and was found to decrease as The particle index (Ia) is based on the concept
the material changes from an angular to a round- that the shape, angularity, and surface texture of a
ed material. Also, C decreased as the amount of uniformly (single) sized aggregate affects not
rounded gravel in a mix increased. only the void ratio but also the rate at which the
Gupta (1985) also reported that the size of sim- voids change when the aggregate is compacted in
ilarly shaped material had an effect on the per- a standard mold (Huang 1962). Through experi-
centage of voids. For all of the materials tested, he mentation with different types of molds and dif-
found that the percentage of voids decreased as ferent types of aggregates, and with a specific
the size of aggregate increased. For example, he procedure, Huang (1962) found a linear relation-
reported a 5% decrease in porosity when the ship between the number of compactive strokes
mean size of rounded gravel increased from 5 (10 to 50) and the void ratio. There was no general
mm to 32 mm. Likewise, a 3% decrease in poros- relationship found with fewer than 10 compac-
ity was reported for a similar change in aggregate tive strokes. The resulting change in the void
size for angular limestone. Therefore, the size of ratio became insignificant when more than 50
the particles has an effect on the AN value. When compactive strokes were used (Huang 1962). The
using AN to characterize angularity it is impor- hypothetical void ratio at zero strokes was line-
tant to ensure that the comparison is done on arly extrapolated from the measured void ratio at
material of the same size. the end of 50 strokes and the rate of void change

10
between 10 and 50 strokes. It was found to be de- 3 represent aggregates passing the 19-mm and re-
pendent on the shape, angularity, and texture of tained on the 12.7-mm sieve; passing the 12.7-mm
the aggregates. Huang (1962) found that this "zero and retained on the 9.5-mm sieve; passing the 9.5-
stroke void ratio" became smaller as the aggregate mm and retained on the No. 4 (4.75-mm) sieve, re-
became more spherical, rounded, and smooth. He spectively. Shapes, a, b, and c represent bulky,
also found that this was more pronounced when elongated, and flat particles, respectively. Gravel
the mold was rhombohedron shaped. (ii) was visually classified as gravel with smooth
The particle index test initially was developed surfaces and rounded corners. Gravel (iii) was
for three aggregate sizes: passing the 19-mm and classified as gravel with rough surfaces and sharp
retained on the 12.7-mm sieve, passing the 12.7- corners and edges. Crushed gravel (p) was made
mm and retained on the 9.5-mm sieve, and pass- up of gravel aggregates as received from the plant.
ing the 9.5-mm and retained on the No. 4 (4.75- No screening between crushed and uncrushed
mm) sieve. For each size, the test involves tamp- gravel was made. Crushed gravel (pp) was made
ing the uniform-sized aggregate into a mold in up of those that were sorted to have at least one
three equal layers using a standard tamping rod crushed face. The shapes, angularity, and rough-
with 10 strokes per layer. The tamping rod is ness classifications were based on visual observa-
raised to a height of 50 mm from the top of the tions by several operators and were not based on
aggregate surface. At the end of the third layer, any measurements (Huang 1962). The results indi-
material is added to make sure that the aggregate cated that the test method is capable of distin-
surface is flush with the rim of the molds. The test guishing the difference between smooth, rounded
is repeated using 50 strokes. The percentage of and rough, angular aggregates by the increasing
voids in the aggregate is calculated using the fol- particle index from gravel to crushed stone. It also
lowing equation: appears that this index is capable of distinguish-
ing different levels of crushed gravel. A modified
version of this test has been standardized by
^n = 1- 100 ASTM. The test is designated as ASTM D 3398-93
S*v
(1996), Standard Test Method for Index of Aggregate
where Vn = percentage of voids at n strokes per Particle Shape and Texture (1996), and it covers both
layer coarse and fine aggregates. For the coarse portion,
Wn = net weight of aggregate in the mold at the maximum size has been increased to 38 mm.
n strokes per layer (g) Depending on the size of the aggregates, different
s = bulk density of the aggregate (g/cc) size molds are used, as shown in Table 10. The
v = volume of mold (cc). mold as prescribed by this standard is cylindrical
in shape. Although the effect of this test method
The particle index (I„) is calculated using the for-
was more pronounced in a rhombohedron-shaped
mula
mold, it could also be seen to a lesser degree with a
circular mold.
Ia = 1.25Vio-0.25Vso-32
The problem with this test method, with respect
where Vio = percentage of voids in aggregates at to coarse aggregates, is in the determination of the
10 strokes per layer volume of voids. Initially, the voids were calculat-
V50 = percentage of voids in aggregates at ed based on the volume of the mold. This required
50 strokes per layer. making certain that the aggregates were filled to
the rim of the mold and the aggregates' surfaces
Typical particle index results reported by were flush with the rim. Because of this it was
Huang (1962) are shown in Table 9. Sizes 1,2, and found that the results were operator dependent.

Table 9. Results of particle index tests for various aggregates. (After Huang 1962.)
Size 1 Size 2 Size 3
Sample Shape a Shapeb Shapec Shape a Shapeb Shapec Shape a Shapeb Shapec

Gravel (ii) 7.5 8.7 9.8 7.1 9.1 9.4 7.5 8.5 —
Gravel (iii) 8.8 10.0 13.2 9.0 10.2 14.0 9.1 10.4 12.2
Crushed gravel (p) 9.5 11.7 13.6 — — — — — —
Crushed gravel (pp) 11.3 14.6 14.8 — — — — — —
Crushed stone 13.4 16.0 16.9 12.4 15.9 17.4 12.6 14.9 16.4

11
 Table 10. Mold size and dimensions of tamping rods for the particle index test. (After ASTM
D 3398-93.)
Aggregate
Aggregate size specimen Mold Mold Rod Rod Mass of
(mm) Mold/Rod size diameter height diameter length rod
Passing Retained designation (kg) (mm) (mm) (mm) (mm) <s>
38.1 25.4 A 13.6 203.2 + 0.2 237.0 ±0.2 21.2 ±0.2 814 ± 0.2 2204 ±10
25.4 19.0 B 13.0 152.4 ±0.2 177.8 ± 0.2 15.9 ± 0.2 610 ± 0.2 930 ± 10
19.0 12.7 B 13.0 152.4 ±0.2 177.8 + 0.2 15.9 ± 0.2 610 + 0.2 930 ±10
12.7 9.5 C 4.0 101.6 ±0.2 118.5 ± 0.2 10.6 ± 0.2 406.9 ± 0.2 276 ±3
9.5 4.75 C 4.0 101.6 ±0.2 118.5 ± 0.2 10.6 + 0.2 406.9 ±0.2 276 ±3

Huang (1965) modified the test method to Bindra and Al-Sanad (1983) conducted tests on
measure the actual volume of the compacted aggregates that ranged from natural aggregates to
aggregate mass. This involved a special device blast furnace slags. They also looked at the effect of
called a volumeter, which fitted exactly over the different gradations (open graded to dense graded)
mold. The volumeter was made up of a bottom and concluded from their study that gradation did
thin flexible membrane and a standpipe. The pro- not affect the weighted particle index value. It
cedure for compacting the sample remained the should be noted that the gradation study included
same, except that at the end of the last layer, a both the coarse and fine fractions. The weighted
flexible membrane was placed on top of the sur- particle index for a given gradation is the weighted
face and water was poured through the stand- mean of the percent retained in each sieve multi-
pipe to a prescribed height. The idea here is that plied by the particle index for that size. The proce-
when water is introduced on top of the mem- dure for determining the weighted particle index is
brane, the thin membrane will deform to match presented in ASTM D 3398 (1996). Bindra and Al-
the surface. Knowing the volume of the mold and Sanad (1983) also concluded from their study that
the volume of water in the volumeter, the actual the particle index for smooth-surfaced, rounded
volume of the aggregate mass is calculated. Using aggregates averaged around 6.5, and for crushed,
this method for volume measurement, more re- high angularity, rough-textured aggregates, the av-
producible results were obtained (Huang 1965). erage was around 17.5.
In addition to modifying the method for measur- They also proposed a new method for calculat-
ing the volume of the compacted aggregates, ing the volume of the aggregate mass in the mold.
Huang modified the test to determine particle in- Instead of using the volumeter device developed
dex of fine aggregates (passing #4 and retained by Huang (1965), they proposed a sand replace-
on the #200 sieve). ment technique. This method involves placement
Bindra and Al-Sanad (1983) modified the par- of a thin flexible membrane on top of the aggregate
ticle index test to test coarse aggregates up to 50- surface. Sand (passing the #25 and retained on the
mm maximum size in their study. In this case, #50 sieve) is then poured onto the surface of the
coarse aggregates are designated as any material membrane to the top of the mold. The net weight of
larger than 2.36 mm (#8 sieve). The test method is sand required to fill to the rim of the mold is ob-
similar to the procedure developed by Huang tained and the volume to fill the mold is obtained.
(1965) with several exceptions. The mold used is The volume of the compacted aggregate mass is the
cylindrical in shape and the dimensions and difference between the volume of the mold and the
weight of the mold and tamping rod sizes are dif- volume of sand required to fill the mold. This
ferent. Details on the dimensions and weights of method appears to be easier and less cumbersome
the mold and tamping rods are shown in Table 11. than the volumeter method.

Table 11. Dimensions of mold and tamping rods. (After

Bindra and Al-Sanad 1983.)
Inside Inside Diameter of Length of Weight of
Aggregate diameter height tamping tamping tamping
size of mold of mold rod rod rod
(mm) (mm) (mm) (mm) (mm) (ft)
20-50 250.0 291.6 26.00 1045 4120
2.36-20 152.4 177.8 15.88 610 930

12
 Macro-Surface Voids
Specific rugosity index
Tons and Goetz (1968) developed the
packing volume concept to characterize the
shape, angularity, and roughness of the ag-
gregates used in bituminous mixtures. The
test was developed for both the coarse (12.7-
mm max) and fine fractions. They decided
that the shape of the particle could possibly Packing
be quantified as a separate value; however, Porosity
Voids
it was difficult to separate the interaction of
angularity and roughness on aggregate per-
formance. They proposed that the effect of
both angularity and roughness be combined
and considered in one term, "rugosity."
The assumption is that the volume of an
individual particle in an aggregate mass de- Packing Volume Membrane

termines the density and voids in the bulk. Figure 13. Packing volume, packing porosity, and geometric
This volume, called the "packing volume," irregularity of aggregate particles. (After Ishai and Tons 1977.)
had to account for the volume of the solid
particle, the volume of internal voids, and the vol-
ume of "outside voids." The "outside voids" vol- The apparent specific gravity of the aggregates
ume is made up of the volume of the dips and can be calculated using ASTM Test for Specific
valleys on the aggregate surface (Fig. 13). The Gravity and Absorption of Coarse Aggregate (C
packing volume can be imagined as a membrane 127). Gpx is determined from the pouring test
around the aggregate. developed by Ishai and Tons (1977). The test in-
The packing volume membrane divides the volves taking two one-sized particles and pour-
voids into two components, the interparticle voids ing them into a standard container using a stan-
and particle surface voids (Ishai and Tons 1977) dard procedure. One of the particles is used as a
(Fig. 13). For one-sized particles, it is assumed that standard (smooth, spherical glass beads) with a
the interparticle voids are constant and this poros- known packing specific gravity, Gps. The other is
ity is the same as that obtained from same one- the test particle for which Gpx is sought. Gpx is a
sized smooth spherical particles. Tons and Goetz function of the ratio of the weight of the test parti-
(1968) assumed that aggregate shape can be math- cle to the standard particle:
ematically defined as ellipsoids. They found that
the porosity calculated from ellipsoids or spheres
of the same size had the same amount of voids. J £Wx J
px ps
Any difference, then, between the porosity of the
smooth spherical particles and the aggregates is
due to the irregularities of the aggregates. where Ws = weight of the standard
Ishai and Tons (1971) developed a specific rug- Wx = weight of the test material
osity (Srv) index to express the total geometric ir- Gp S = packing specific gravity of standard
regularity of the particle. Srv will be approximately material
equal to zero for smooth, spherical particles: Gp X = packing specific gravity of test mate-
rial.
\ ft J X
V„ A schematic of the pouring test is given in Fig-
Sw = 100 = 100 1- P
Vn J ap ure 14. The specifications for the pouring test ap-
paratus for the coarse aggregate fraction are giv-
where Srv = specific rugosity (%) en in Table 12.
Vsr = volume between the packing volume Some typical results as presented by Ishai and
membrane and the volume of macro Tons (1977) are shown in Figure 15. The specific
and micro surface voids rugosity of two sizes, 12.7 to 15.9 mm and #3 to
Vp = packing volume of the particle #4 sizes for natural gravel, crushed gravel, lime-
GpX = packing specific gravity stone, and beach pebbles are plotted. The specific
Gap = apparent specific gravity. rugosity increases as the material becomes more

13
 Table 12. Specific setups for coarse fractions
Supports for the pouring test. (After Ishai and Tons
1977.)
Dimensions Coarse frat tions
(cm) 12.7 to 15.9 mm #3 to #4
Bin diameter 16.0 16.0
Funnel orifice diameter 3.0 1.44
Aggregate head 12.0 12.0
Pouring height 21.0 21.0
Container diameter 12.2 10.3
Container height 15.2 11.8
Bead diameter (avg.) 16.0 6.0

volves allowing the fine aggregates to freely fall

into a calibrated cylinder from a specified height.
The exception here is that the aggregates are not
separated into various sieve sizes, as is done with
the other index tests. The test for fine aggregates
is designated by ASTM as Test Method C 1252.
The weight of aggregate and the bulk specific
gravity are used to calculate the uncompacted
void content:
D = Bin Diameter H Pouring Height
a = Funnel Office Diameter $ Container Diameter M
c = Bin Height h Container Height Vcyl'
r
b = Aggregate Head
uvc = - xlOO
Figure 14. Schematic description of the pouring device. "'cyl
(After Ishai and Tons 1977.)
where UCV = uncompacted void content (%)
V„I = volume of cylinder (cm3)
angular and rougher. Also shown in the figure is
M = mass of aggregate in cylinder (g)
the effect of aggregate size on the specific rugos-
GSD = bulk specific gravity of aggregates.
ity. The value of the specific rugosity increases as
the size of the material decreases. UCV increases with increasing angularity and
roughness of the aggregates. Aldrich (1996) mod-
Uncompacted voids in aggregates ified this test for coarse aggregates. A schematic
This test method is similar to the specific rug- of his test apparatus is shown in Figure 16. The
osity test. Initially developed for fine aggregates coarse aggregate for the test passed the 19-mm
by the National Aggregate Association, it in- sieve and was retained on the #4 sieve. The test
25
I I I
D Natural Gravel (#3 to #4)
o Crushed Gravel (#3 to #4)
20 — A Limestone (#3 to #4) —
O Beach Pebbles (#3 to #4)

m On
. 15

o> ±A
••
10 —
- ■ —

a ■
c/J ■ Natural Gravel (1/2 to 3/8 in.)
5 — • Crushed Gravel (1/2 to 3/8 in.)
▲ Limestone (1/2 to 3/8 in.)
♦ Beach Pebbles (1/2 to 3/8 in.)

I I I
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
Packing Specific Gravity (Pouring), Gpx
Figure 15. Packing specific gravity and rugosity for different aggregates.

14
 method for coarse aggregates was developed by
the French in 1981. The basis of this test is that the
4" flow rate of an aggregate mass is affected by an-
gularity, surface texture, and the bulk specific
gravity of the aggregate. The equipment neces-
\$60 i sary for this type of test method can be found in
the Quebec Ministry of Transportation Aggre-
H h
0.5 4.5" gates Laboratory in Quebec City, Quebec. Figure
L 17 shows the test apparatus.
The flow coefficient (Ce) of an aggregate is the
time required in seconds for 7 kg of material to
Fine Aggregate Coarse Aggregate flow through a specified opening under a vibra-
Figure 16. Test apparatus for ASTM C1252 and mod- tory frequency of 50 Hz. The opening is 60 mm
ified ASTM C 1252. (After Aldrich 1996.) for material passing the 20-mm sieve and re-
tained on the 4-mm sieve. The maximum size cur-
apparatus, as pointed out by Aldrich (1996), is rently tested with this equipment is 20 mm. The
similar to that developed by Ishai and Tons (1977) flow coefficient (Ce) is determined from the fol-
for the Specific Rugosity Tests. The test method is lowing equation:
similar to that used for fine aggregates and the
uncompacted void content is calculated using the n fxGsb
same equation used for the fine aggregates. Aid-
rich (1996) reported that the UCV increased with
where Ce = flow coefficient
increasing angularity of the aggregates.
t= time (sec)
Gsb = bulk specific gravity
Time index k= constant (1 second).
Several test methods exist for fine aggregates
where the rate of flow through a standardized or- The test involves first drying 10 kg of aggre-
ifice is used to index the material shape, angular- gate. This amount is then placed into the sample
ity, and texture (Rex and Peck 1956, Tobin 1978). tube. The vibrating table is turned on and 1 kg of
The time of flow is usually compared to a stan- material is allowed to collect in the pan on the
dard material, such as Ottawa sand. A similar test floor. The pan is set on a scale and the weight of
the aggregate in the pan is monitored. After 1 kg
of material has collected in the pan, a stopwatch
is turned on. The watch is stopped when an addi-
tional 7 kg of material has collected in the pan.
The amount of time for 7 kg to collect in the pan is
noted. The test is repeated five times and the
average time value is determined.
The same material used for the image analysis
was tested using this equipment. The results are
shown in Table 13. The bulk specific gravity of the
aggregates was assumed to be 2.6.
The results indicate that, based on the flow
coefficients, crushed stone is more angular than
crushed gravel. This test can be conducted
quickly; further work in this area is warranted.

Table 13. Average flow coefficients for

Vermont subbase materials.
Average time Flow
Material (s) coefficient
Crushed gravel 33.45 92.2
Crushed stone 44.13 114.7
Figure 17. QMOT time index test apparatus. Pike crushed stone 44.62 116.0

15
LABORATORY RESPONSE OF (Thompson and Smith 1990). They reported the
ANGULAR MATERIAL range to be about 200 to 240 MPa under a bulk
stress of 138 kPa. Ishai and Gelber (1982) also
From static triaxial tests, it has been shown that found similar trends for bituminous mixes. They
rounded aggregate materials produce significant- suggested that this was to be expected because in
ly higher permanent deformation than angular the resilient modulus test, the applied stress levels
aggregate soils. This is reflected in the angle of in- are very low and only with near-failure conditions
ternal friction (<j>). During compression, it is com- will the effect of the geometric irregularities be ev-
monly found that rounded particles are able to ident; for example, in terms of particle interlock-
slip easily, whereas angular materials have to ing. However, the deformation under the same
overcome the higher frictional forces at the contact applied stress level, when the base course is satu-
interfaces. Generally, § increases with increasing rated, will be high, and the angularity and rough-
angularity (Holtz and Kovacs 1981). Holtz and ness of the aggregates will have a greater effect on
Kovacs (1981) also found that as the surface pavement performance.
roughness increased, so did (j). The effects on (j) by There is very little information in the literature
angularity, surface roughness, and other factors on the performance of base course layers as a func-
are shown in Table 14. tion of the shape, angularity, and roughness of the
aggregates. The results of only two studies on the
Table 14. Summary of factors affecting <)>. (After effect of aggregate geometry and surface rough-
Holtz and Kovacs 1981.) ness were found in the literature. One was by Hol-
Factors Effect ubec and Wilson (1970) and the other by Barksdale
and Itani (1994). Holubec and Wilson (1970) used
Void ratio (e) eltyl
Aiyr crushed gravel and crushed stone in their study.
Angularity (A)
Grain size distribution QV
They looked at the effects of angularity and the
Kt(t>t
proportion of crushed material on base course per-
Surface roughness (R)
Moisture content (w) wtfyl slightly
formance. The samples were compacted at an
No effect (with constant e)
optimum moisture content of 4.5% using a 4.5-kg
Particle size (S)
Intermediate principal stress
hammer dropped 457 mm and applying 25 blows
V-^tx per layer. The test samples were 102 mm in diame-
Overconsolidation or prestress Little effect
ter and 203 mm in height. The maximum aggre-
gate size was 9.5 mm.
The information in Table 14 relates to static Holubec and Wilson (1970) report in terms of
loading conditions. However, the loading condi- the cyclic creep strain. The definition of cyclic
tion on a pavement structure is not static, but cy- creep strain is unclear in the report and is taken by
clic. Also, the load levels applied are usually not this author to refer to the total strain minus the
near the shear strength of the material except elastic strain. They found that the cyclic creep
when the material is saturated, as is the case dur- strain after 5000 load repetitions decreased as the
ing spring thaw. The data in Table 14 may or may percent of crushed particles increased (Fig. 18).
not apply to pavement structures. Very little data The samples were made by blending different
are published on the performance of base course amounts of crushed aggregate with the parent
materials under cyclic loading. This is probably rounded gravel aggregates. With respect to sphe-
due to the erroneous assumption that failure of ricity and angularity, they reported that, as the ag-
base course materials does not occur. gregates became angular, the cyclic creep strain,
With the advent of mechanistic pavement de- after 5000 load repetitions, increased (Fig. 19). The
sign procedures, several studies have been pre- results appeared to be contradictory, as one would
sented on the resilient modulus of base materials. expect the creep strains to decrease with increas-
Base materials have been identified as stress de- ing angularity. The results may be explained by
pendent and several models exist for predicting the fact that there is a larger contact area with
the resilient modulus of base materials, such as rounded material than with angular material.
the Ka-K6 model in SUPERPAVE (Lytton et. al Thus it will require a larger shear force at the
1993) and the "universal" Ki-K3 model (Witczak points of contact to move the aggregates. This
and Uzan 1988). However, the resilient moduli re- could also explain why the resilient modulus re-
ported in the literature for natural gravel and ported in the literature for rounded and angular
crushed rock were quite similar to one another materials are similar. This is only speculation and

16
 3.00

0 25 50 75 100
Percent of Crushed Particles

Figure 18. Effect of crushed particles on cyclic creep strain. (After Holubec
and Wilson 1970.)
4.00
I I I I I I I I
- o

I
8 3.00

I
&
o
•o - -
'^cr^o
IS
o

5
o 2.00
in
o
o
<
Q.
0) 0 —
2> 1.00 -
O
o
o -
>-
Ü

0.00 I I I I. I I I I
1.00 0.90 0.80 0.70 0.60
Rounded Angular
Sphericity and Angularity
Figure 19. Effect of combined sphericity and angularity on cyclic creep of
base course materials. (After Holubec and Wilson 1970.)
further work in this area may be required. In the The other work was done by Barksdale and
same study, the friction angle was found to Itani (1994). They presented results of the effect of
increase with increasing angularity (Fig. 20). aggregate shape on the resilient and permanent
Based on their study, Holubec and Wilson recom- deformation behavior of five base course materi-
mended that repeated load tests be conducted on als. The gradation used in the study is shown in
characterizing the performance of base course Table 15. The medium gradation in Table 15 is
materials and cautioned on the use of the angle of similar to the VAOT fine gradation for subbase of
internal friction as a parameter for predicting crushed gravel.
base course performance. In my opinion, with
thawing soils, the friction angle may be a good Table 15. Gradation of test aggregates. (After
indication of the ability of the base course to resist Barksdale and Itani 1994.)
permanent deformation. When the base is satu- Percent passing
rated, the load levels commonly applied on pave- Gradation 38.2 mm 19 mm 9.5 mm No. 4 No. 40 No. 200
ment structures when they are dry may be high Medium 100 80 60 45 13 4
enough to move the base response from the elas- Coarse 100 65 43 27 7 0
Fine 100 85 70 58 25 10
tic to the plastic (failure) zone.

17
 45 i
I I I I ' I I

CD
44
CD
D .^o
-©-
c
o
43 — —
o
\_
LL
«
E 42
a>
1: a/^
"5
CD
g> 41 —
<

40 I i I i i i I I
1.00 0.90 0.80 0.70 0.60
Rounded Angular
Sphericity and Angularity

Figure 20. Effect of sphericity and angularity on the angle of internal friction.
(After Holubec and Wilson 1970.)

■ Granite
9 Gravel
A Shale
^ Quartzite
Y Limestone

0.0 0.2 0.4 0.6 0.8

Short/Intermediate = p
Figure 21. Shape classification of aggregates. (After Barksdale and Itani 1994.)

The test materials, shape, and surface charac- men diameter be at least six to seven times the
teristics as reported by Barksdale and Itani (1994) maximum particle size, to minimize the bound-
are shown in Figure 21. The samples were com- ary effects by the large aggregates. Scalping of
pacted using a vibratory compactor at the opti- large aggregates and replacing equivalent
mum moisture content. The test specimens were amounts of smaller aggregates changes the over-
152 mm in diameter and 305 mm in height. These all structure of the material. Dawson* reported
dimensions are acceptable provided the maxi- that in unreported experiments at the University
mum size of the aggregates is no more than 25 of Nottingham, similar results were obtained on
mm. Sweere (1990) found that testing aggregates dry specimens in which the grading curve was
containing 51-mm particles in a 152-mm-diameter
mold produced inconsistent results. He recom- * Personal communication, Andrew Dawson, University of
mended that, for large particle sizes, the speci- Nottingham, United Kingdom, 1992.

18
either translated laterally or truncated. This in Table 16. AIF values
turn produces resilient and plastic responses that of test aggregates.
may not be representative of the actual material. Base material AIF
If there is a sufficient amount of large aggregates
Granite gneiss 500
(greater than 10%), large-scale triaxial testing is Limestone 50
recommended. Besides aggregate geometry, Shale 1675
Barksdale and Itani (1994) looked at the influence Quartzite 150
of the fines, the plasticity of the fines, gradation, River gravel 3700
and moisture content on base course perfor-
mance. This AIF value is specific to the aggregates tested
To quantify the effect of the aggregate geome- by Barksdale and Itani (1994) and is not recom-
try and surface roughness, Barksdale and Itani mended for other aggregates. If this approach is
(1994) developed the aggregate influence factor taken to characterize base course performance,
(AIF), which was based on a multiple regression new AIF values will have to be developed for oth-
between observed laboratory performance and er aggregates. The AIF values for the test aggre-
various indices that described the aggregate gates are presented in Table 16. The AIF values
shape and texture. For the aggregates tested, they decrease with increasing angularity and rough-
found that the AIF was a function of the spherici- ness.
ty, roundness, surface roughness, and angularity The resilient modulus test was conducted using
of the aggregate. They employed methods used the procedure prescribed in AASHTO T-274-82.
by geologists to quantify sphericity and round- For permanent deformation studies, the speci-
ness as described earlier in this report. For surface mens were subjected to 41 kPa confining pressure
roughness, Barksdale and Itani (1994) used a and subjected to 70,000 load repetitions at a prin-
scale from 0 to 1000 to quantify glassy to very cipal stress ratio of four or six. Barksdale found
rough particles, respectively. They then rated the the resilient modulus of rough angular materials
roughness of the aggregates based on visual ex- to be higher than the rounded material by about
amination. For example, for the aggregates in the 50% at low bulk stress values and about 25% at
study, the AIF was as follows: high bulk stress values (Fig. 22). This is much
higher than the values reported by Thompson
AIF = 2500* (v + R) - (A + SR) and Smith (1990). There appears to be some in-
where AIF = aggregate angularity factor consistency in the results found in the literature.
\|/ = average sphericity value The effect of moisture was also pronounced.
R= average roundness value Barksdale and Itani (1994) reported that under
SR = surface roughness coefficient. drained conditions, the resilient modulus

100,000

Granite Gneiss (Yd=141 pcf, w=6.2%)

Gravel (yd=126 pcf, w=3%)
Shale (Yd=130 pcf, w=4%)

a
DC
10,000

5000
10 100
Sum of Principal Stresses (psi)

Figure 22. Influence of material type and state of stress on resilient modulus.
(After Barksdale and Itani 1994.)

19
 0.0
10 12
Percent of Fines

Figure 23. Influence of material type, stress level, and percent fines on plas-
tic strain. (After Barksdale and ltani 1994.)

decreased by 50% (compared to the optimum flow through a known orifice. This method is
moisture content) for river gravel at a bulk stress fairly simple and has been used for fine aggre-
of 103 kPa. For undrained conditions, this reduc- gates; it can be modified for coarse aggregates up
tion will be much higher. He also reported that to 50 mm in diameter.
the amount of fines in the base course materials Several research studies have shown that the
had a significant effect on the amount of plastic shape, angularity, and roughness have a signifi-
strain (Fig. 23). cant effect on base performance. Some of these
studies have shown that there can be up to 50%
change in resilient modulus of base materials due
SUMMARY AND CONCLUSIONS
to geometric irregularities. It is not conclusive
There are several methods for quantifying the whether the resilient modulus is a good indica-
shape, angularity, and roughness of base course tion of the effects due to changes in shape, angu-
aggregates. Methods developed by geologists can larity, and roughness of the aggregates. However,
be classified as direct methods and involve actual what is clear in the literature is that moisture and
measurements on the aggregates. Basic measure- amount of fines have a significant effect on base
ments of lengths are used to quantify the shape of course performance. Increasing moisture content,
aggregates. The angularity and roughness can as during the spring thaw, can increase the failure
also be determined with complex measurements, rate of the base course. Further study is needed to
but several charts have been developed for sim- consider the combined effect of aggregate irregu-
plifying the process. These methods are more in- larities, moisture, and fines content on base
volved and time consuming. However, they can course performance.
be considered as standards.
Indirect methods for quantifying the shape, RECOMMENDATIONS
angularity, and roughness have been developed.
The shape, angularity, and roughness are usually The results of this review indicate that for base
combined together as it is fairly difficult to sepa- course materials with maximum sizes ranging
rate the effects of the individual components. between 25 mm and 50 mm, the particle index
Most indirect measurements are based on the test is best suited for indexing the angularity of
change in the voids of a single-sized bulk mass of the material. The term "angularity" as used in
aggregates. The tests are similar to one another this recommendation includes the shape and sur-
and involve compaction of the aggregate mass in face characteristics of the aggregate. One of the
a given level. Some of the indices developed are drawbacks of using this test is that it is time con-
the angularity number, particle index, and rugos- suming.
ity. Another index test involves the amount of VAOT requirements state that any test method
time required for a given aggregate gradation to used for characterizing the angularity of base

20
course materials be simple and quick to use. aggregate shape on base behavior. Transportation
Either the flow coefficient or the uncompacted Research Record 1227, p. 171-182.
void tests have the greatest potential to meet the Bikerman, J.J. (1964) Adhesion of asphalt to stone.
above requirements. Currently, these tests are MIT Civil Engineering, Research Report R64-3.
limited to aggregates with a maximum size of 20 Bindra, S.P., and H.A. Al-Sanad (1983) Particle
mm, and modifications to both the equipment index evaluation of aggregates for unbound pav-
and test procedures are required. It is recom- ing mixtures. In Proceedings of the Fourth Confer-
mended that a research program be conducted to ence of Road Engineering, 22-26 August. Jakarta,
develop an index test for angularity of aggregates Indonesia: Road Engineering Association of Asia
to a maximum size of 50 mm using both the flow and Australasia, p. 85-93.
coefficient and the uncompacted void tests. The British Standards Institution (1975) BS 812: Test-
results from both these tests will be correlated ing Aggregates.
with the particle index, because this index has Gupta, R.D. (1985) Effect of size of particles on
been developed for larger size aggregates. their angularity. Journal of the Institution of Engi-
It is also advisable, but optional, that the neers (India), 65(CI 5): 210-214.
shape, angularity, and surface texture of the Holubec, I., and K.H. Wilson (1970) A cyclic creep
aggregates be characterized using the techniques study of pavement materials. Final Report, De-
developed in the geological arena. These will be partment of Civil Engineering, University of
used to calibrate the index results obtained from Waterloo, Ontario, D.H.O. Report No. RR163.
the flow coefficient and the uncompacted void Holtz, R.D., and W.D. Kovacs (1981) An Introduc-
tests using a similar method developed by Barks- tion to Geotechnical Engineering. Englewood Cliffs,
dale and Itani (1994). The geological characteriza- New Jersey: Prentice-Hall, Inc.
tion can also be used to correlate resilient modu- Huang, E.Y. (1962) A test for evaluating the geo-
lus and shear strength test results using a similar metric characteristics of coarse aggregate parti-
approach used by Barksdale and Itani (1994) cles. Proceedings, American Society for Testing and
again. Materials, 62:1223-1242.
The index angularity measurements need to be Huang, E.Y. (1965) An improved particle index
tied to the resilient and strength properties of the test for the evaluation of geometric characteristics
aggregates. Ideally, the resilient modulus and of aggregate. Michigan Highway Research Pro-
shear strength tests should be conducted under ject No. 86546.
optimum moisture and saturated conditions. The Ishai, I., and H. Gelber (1982) Effect of geometric
critical time for the base courses in seasonal frost irregularity of aggregates on the properties and
areas is during spring thaw. Therefore as a mini- behavior of bituminous concrete. Proceedings, As-
mum it is recommended that the tests be con- sociation of Asphalt Paving Technologists, 51: 494-
ducted on saturated samples under undrained 521.
conditions. As a corollary to the testing program, Ishai, I., and E. Tons (1971) Aggregate factors in
moisture density relationships will be developed bituminous mixture designs. University of Mich-
using the large-scale Proctor testing device. These igan, Ann Arbor, Report 335140-1-F.
relationships will be compared with those Ishai, I., and E. Tons (1977) Concept and test
obtained from standard testing procedures used method for a unified characterization of the geo-
atVAOT. metric irregularity of aggregate particles. Journal
of Testing and Evaluation, 5(1): 3-15.
Krumbein, W.C. (1941) Measurement and geolog-
LITERATURE CITED
ical significance of shape and roundness of sedi-
Aldrich, R.C. (1996) Influence of aggregate prop- mentary particles. Journal of Sedimentary Petrology,
erties on heavy duty pavements. Transportation 11(2): 64-72.
Research Board. Lees, G. (1964) The measurement of particle
ASTM D 2488-90 (1996) Standard practice for de- shape and its influence in engineering materials.
scription and identification of soils (visual-manual Journal of the British Granite and Whinestone Federa-
procedure). ASTM, vol. 04.08, Soil and Rock. tion, London, 4(2): 1-22.
ASTM D 3398-93 (1996) Standard test method for Lytton, R.L., J. Uzan, E.G. Fernando, R. Roque, D.
index of aggregate particle shape and texture. Hiltunen, and S.M. Stoffel (1993) Development
ASTM, vol. 04.03, Road and Paving Materials. and validation of performance prediction models
Barksdale, R.D., and S.Y. Itani (1994) Influence of and specifications for asphalt binders and paving

21
mixes. Federal Highway Administration, Turner- Tobin, R.E. (1978) Flow cone sand tests. ACI Jour-
Fairbanks Highway Research Center, Strategic nal, p. 1-12.
Highway Research Program Report PA 357. Tons E., and W.H. Goetz (1968) Packing volume
Rex, H.M., and R.A. Peck (1956) A laboratory test concepts for aggregates. Highway Research Record,
to evaluate the shape and surface texture of fine no. 236, p. 76-96.
aggregate particles. Public Roads, 29:118-120. Wadell, H. (1932) Volume, shape and roundness
Shergold, F.A. (1953) The percentage of voids in of rock particles. Journal of Geology, 40: 443^451.
compacted gravel as a measure of its angularity. Wilson, J.D., L.D. Klotz, and C. Nagaraj (1995) Auto-
Magazine of Concrete Research, 5(13): 3-10. mated measurement of aggregate indices of shape.
Sweere, G.T.H. (1990) Unbound granular bases Federal Highway Administration Report FHWA-
for roads. Ph.D. Thesis, University of Delft. RD-95-116.
Terzaghi, K., and R.B. Peck (1967) Soil Mechanics in Witczak, M.W., and J. Uzan (1988) The universal
Engineering Practice. New York: John Wiley and airport pavement design system, Report I of rV:
Sons, Second edition. Granular material characterization. University of
Thompson, M.R., and K.L. Smith (1990) Repeated Maryland, College Park, Maryland.
triaxial characterization of granular bases. Trans- Wright, P.J.F. (1955) A method of measuring the
portation Research Record, 1278: 7-17. surface texture of aggregate. Magazine of Concrete
Research, 7(21): 151-160.

22
 Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestion for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington,
VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503.
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED
January 1998
4. TITLE AND SUBTITLE 5. FUNDING NUMBERS

Quantification of Shape, Angularity, and Surface Texture

of Base Course Materials
6. AUTHORS

Vincent Janoo

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

REPORT NUMBER
U.S. Army Cold Regions Research and Engineering Laboratory
72 Lyme Road Special Report 98-1
Hanover, New Hampshire 03755-1290

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING

AGENCY REPORT NUMBER
State of Vermont Agency of Transportation
133 State Street
Montpelier, Vermont 05602

11. SUPPLEMENTARY NOTES For conversion of SI units to non-SI units of measurement consult Standard Practice for Use of the
International System of Units (SI), ASTM Standard E380-93, published by the American Society for Testing and Materials,
1916 Race St., Philadelphia, Pa. 19103.
12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Approved for public release; distribution is unlimited.

Available from NTIS, Springfield, Virginia 22161.

13. ABSTRACT (Max/mum 200 words)

A state-of-the-art review was conducted to determine existing test methods for characterizing the shape, angular-
ity, and surface texture of coarse aggregates. The review found direct methods used by geologists to determine
these characteristics. These methods involve physical measurements of individual aggregates and are very labori-
ous and time consuming. Engineers have developed index tests (indirect methods) to quantify the combined
effect of the shape, angularity, and surface texture of coarse aggregates in terms of changes in the voids in the
aggregate bulk. A description of both the direct and indirect methods is provided in the report. Also, the effect of
shape, angularity, and surface texture of coarse aggregates on the base course performance was reviewed. It was
found that there is some contradiction in the published data on resilient modulus. Shape, angularity, and surface
texture of coarse aggregates clearly influence the angle of internal friction.

14. SUBJECT TERMS 15. NUMBER OF PAGES

Angularity Roughness 28
Coarse aggregates Shape 16. PRICE CODE
Resilient modulus
17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT
OF REPORT OF THIS PAGE OF ABSTRACT

UNCLASSIFIED UNCLASSIFIED UNCLASSIFIED UL

NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. Z39-18
298-102