Nr 39 1978
Statens väg- och trafikinstitut (VI'I) - Fack - 581 01 Linköping
National Road & Traffic Research Institute - Fack - 581 01 Linköping - Sweden
Methods for the Testing of Aggregate Strength
by Peet Höbeda
39 Paper resented at the International Symposium on Aggregates
Nr 39 - 1978 Statens väg- och trafikinstitut (VTI) - Fack - 581 01 Linköping
National Road & Traffic Research Institute - Fack - 581 01 Linköping - Sweden
Methods for the Testing of Aggregate Strength
by Peet Höbeda
39 Paper resented at the International Symposium on Aggregates
Methods for the Testing of Aggregate Strength by Peet Höbeda
The National Swedish Road and Traffic Research Institute (VTI)
Fack
8 581 01 LINKÖPING
The cannetatton between tabonatong methodt Act the detenmtnatton 05 dggnegdte étnength tö theuAAed and the ttandandtzed methodb tn dtäéedent eountnteé cadnetdte netattvetg wett. The Lab Angeteé method, whteh tn the one moat tn ate, hat centatn advantageA and the beAt hepeatabtttty and thoutd be tnetuded tn the heeommenda~ .tton by the RILEM Commtttee GM 26. In addttton, Apeetat anatyéeé may have to be edddted ont, e.g. d wet degndddtton tent ab
aggnegateé wtth tatté aetony tent vatueé may have a tendency to degndde tn the nadd. Å Atmpte etaéététeatton can be tnttnumentat tn gtndtng the agghegateé that nequtne Apeetat anatgAeA.
The strength properties of the aggregates are determined according to a number of standardized methods in diffe-rent countries. The most common methods are the Los
Angeles test and the Impact and Crushing tests. In the
first case, the aggregate is exposed to the stresses of falling steel shots in a rotary drum equipped with a flange, in the latter two cases it is loaded in a con
fined condition in a cylinder. The relevance of these methods, which were developed decades ago, is, however,
dubious, and only a few investigations have been made to correlate results from the laboratory with field
conditions, especially modern road constructions with
low void gradings.
One literature study shows that as a rule there is a correlation between these aggregate test methods (1). A ring analysis that was made in 1975 by the RILEM Committee GM 26 produced a surprisingly good result
study gave a somewhat too idealized picture of the
correlation. In that particular case all the aggregates were 8-11 mm and one of the conditions of correlation
is that aggregates not too unequal in size are compared. The strength is influenced by the flakiness of the
aggregate and different sizes have different particle shapes. Furthermore, in the Los Angeles testing the strength increases with the particle size owing to the fact that the particles can be loaded individually by steel shots. In the two latter testings the results become poorer with an increasing particle size since
the confined mass of particles is loaded in the cylinder, which results in a reduction of the number of particle
contacts. Petrographically based exceptions to the
above-mentioned may, however, be found.
From the literature study (1) it appears that the Los Angeles method has a better repeatability than the Im pact and Crushing strength methods. A ring analysis made between 16 different European laboratories (2)
showed, however, a poor reproducibility, which probably has to be ascribed to differences in the equipment and to the performance of the test.
Since the Los Angeles method is used in several
countries and, as mentioned above, has advantages as a
test method, it should be included in the recommendation
by the RILEM Committee GM 26. It seems unnecessary to have more than One test for aggregate strength. For the time being, the reproducibility of the Los Angeles test leaves much to be desired and hence efforts must be made to improve the method. Some standard materials, for in stance, could be useful for calibration purposes. It is also necessary to standardize a single test size, which presently varies in different countries. The B gradation
(9.5 l9.9 mm) according to ASTM C 131 seems to be the gradation presently most in use and is already included in the Finnish road specifications. As for the B
grada-tion, the test sieves are in agreement with the ISOw
series.
The Impact tests and the Crushing tests are performed in essentially differents ways in different countries, and in the first case the equipment differs very much in design. Hence it is more difficult to include these methods in the recommendation than the Los Angeles
method.
Experience gained in several countries has proved that aggregates can be unsuitable in spite of satisfactory values according to the standardized test methods above. The aggregates are tested in a dry condition according
to standards, but moisture can, to an essential degree,
weaken certain aggregates. Therefore it should be logical to test aggregates in a moist condition. This complicates the tests to a certain degree and presently it seems advisable to perform certain tests for moisture sensivity with problematical aggregates instead. Such tests have been developed e.g. in the USA, "Production of Plastic Fines in Aggregate" (AASHTO T 210) and
"Washington Degradation Test", and in France the Micro deval Method (dry and wet). In the latter case grinding shots are added. In Sweden, as regards base course gra vel, wet degradation on the whole gradation without the fines being removed has been tested in a laboratory mixer of a planetary type. This method has advantages, but it is more complicated to prepare the samples. Thus, in addition to the Los Angeles method, a wet de gradation test should be carried out and as such the
French Microdeval method seems to have certain advantages.
According to this method, it is not necessary to investi
gate all the aggregates, i.e. aggregates of ingeous and metamorphic origins. Therefore it is advisable to employ a simple classification of road materials.
The classification below, which is based on petro-graphical groups, is founded on Swedish experience and is suitable for relatively homogeneous aggregates. 1. "Primary" rocks (granite, gneiss, greenstone etc) 2. "Primary" rocks, rich in mica
3. "Primary" rocks, weathered 4. Limestone
5. Shale 6. Sandstone
7a. Synthetic aggregates, non porous 7b. Synthetic aggregates, porous
If the aggregate is classified as belonging to group 1, it is enough to test e.g. the Los Angeles value, but the result should not be uncritically compared with Los Angeles results from other groups. Other groups require special analyses, often a wet degradation test. It may also be necessary to carry out a weathering test, although in many cases the results of weathering and wet degradation tests seem to be in agreement. The synthetic aggregate group also includes certain waste products, which can be of essential different proper ties. They require special analyses of chemical re
sistance, for instance, and of the occurence of sub-stances dissolvable in water, etc.
For road surfacing aggregates special properties have to be investigated, the polishing resistance, for in
stance, or if the road is exposed to traffic with studded tyres the wear resistance has to be investigated. These properties of the aggregate are not directly related to the strength. The test for wear resistance must not be confused with wet degradation tests previously discussed, but has to be tested on a grinding wheel (e.g. the
British Abrasion Value, BS 812) or according to some other sutiable method. In Sweden promising results have lately been obtained using the so called Tröger appara tus of West-German origin, which has a bundle of
hardened steel pins that beats against a rotary sample body.
References
l. Höbeda, P. Suggestions to the International Standardization of Test Methods for Aggregate Strength. Statens Väg och trafikinstitut, Medde-lande 102, 1978.
Leins, W,, Kohler, G., Meyer, G., Berg, von, P.
Beantspruchungen und die Pr fung von Gesteinsplitten. Internationale Splittpr fverfahren und deren Aussage kraft und die Beeinflussung der Prufwerte durch die Kornform. Strassenbau und Strassenverkehrstechnik. Heft 156, 1974.
G e r man Im pa ct Te st , Zs p, B e n s b e r g
Ge
rm
an
Im
pa
ct
Te
st
,
Za
p,
N
rn
be
rg
30 $ ' Los Angeles, Budapest
X , Liege 0 X . O " - , Paris X . 20'* )( o O 0 )( Xi IQ.. % % % 10 20 30
Los Angeles Test
30 4
. Los Angeles, Budapest
)( "" ,Liege oxx .. 0 ' " , Paris 20--X o
O
Q X
&
10--r
;
. &
10 20 30Los Angeles Test Figure la b . Correlation between methods for aggregate
strength according to RILEM GM 26. Aggregate
% .! Los Angeles, Budapest N * 30 "w ? X w'" _ 9 Lie89 .L.) £3 O - " = Paris E 4 i m 0 .X 0 c n 4 .: U)
£3
.
x
'
L) g 20" m4 00a;
Ox -
' X 10" &( l L l I l r IO 20 30Los Angeles Test
%
30-N , . å )(åå
X
4.3 u2
E H :m 20"' oE
x
3
&
i! 10 an 0 Splittschlag, Bensberg O - " - , Niirnberg i , . L i I I 10 20 30 Druchzertr mmerung Figure Zamb. Correlaticn between methods for aggregatesize 8~ll mm.
& 30!) O : ...g & > m m H
ä
2 X3
__
0°
4 20 X 0 . X10-0Los Angeles, Budapest
X -" , Liege
O - " - , Paris
=
?
%
%
%
%
;
10 20 30 40 50 60 70
Swedish impact test 30" D. U) N .
ä
m6
H 4.) Uä
E 20 '"' O%
.o
E
.
Q) wQ
10 lb C Zsp, N rnberg O Zsp, Bensberg 1 L [ L L l | ' l T T I 10 20 30 40 50 60 Swedish.impact testFigure 3aébt_Correlation between methods for aggregate
strength according to RILEM GM 26.Aggregate size 8 11 mm.
