349
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
1
Some notes concerning the dry density testing standards
Quelques remarques concernant les descriptions relatives aux essais de densité sèche
Imre E.
1, 2
, Lőrincz
J.
2
, Gerendai E.
1
, Szalkai R.
1
1
Ybl Miklós Faculty of Arch. and Civil Eng., Szent István University
2
Department of Geotechnics, Budapest University of Technology and Economics
Lins Y., Schanz T.
Ruhr University, Bochum, Germany
ABSTRACT: The results of two doctoral programs concerning the dry density of sands are as follows: (i) the ratio of the minimum
and the maximum dry density of sands is about constant
, (ii)
the dry density of the soil fractions slightly increases with the maximum
grain diameter
d
max
being the diameter range doubled with increasing grain diameter. These results are used through a statistical
analysis
to show that
the
e
max
test of the German DIN may be biased. The possible reason for the bias is arching due to the too small
ratio of the diameter and the height of the mold. As a by-product of the research, the earlier finding that the maximum density of the
specimens may be obtained indifferently using a Modified Proctor procedure or a vibrating table is extended to the DIN (implying
that the density methods of the calibration chamber sands are basically equivalent to the DIN).
RÉSUMÉ : Deux thèses de doctorat ayant pour objet
l’é
tude de la compacité des sables ont récemment formulé les importantes
conclusions suivantes : (1) le rapport de la compacité minimale et maximale des sables est plus ou moins constant ;
(2) l’étendue de
s
fractions de diamètre croît avec le diamètre des grains, par conséquent la compacité des fractions croît également. En partant de ces
conclusions, on a procédé à la comparaison (partiellement statistique) des valeurs de compacité minimales et maximales de cinq
groupes donnés
. Les résultats ont montré qu’une partie des normes actuelles conduisent
à des erreurs déterminées dans la valeur de
compacité minimale et ne conduisent pas à des erreurs déterminées dans la valeur de compacité maximale
. L’explication
en est,
probablement, l’apparition de l’effet de
voûte aux pots et vases de largeur différente, de hauteur constante..
KEYWORDS: sand, dry density, arching, Proctor, statistical test, coefficient of variation
1 INTRODUCTION
Two doctoral programs on the dry density of sands - where the
soils used were indeed carefully chosen, artificial mixtures of
natural grains (Lőrincz 1986, and Kabai, 1968)
- ended with the
two basic observations.
Kabai (1968 to 1974)
found that the ratio of the minimum
and maximum dry density was basically constant. Its value
started to decrease as the soil became slightly plastic. He stated
this
on the basis of his
experimental studies made on
continuous, artificial mixtures of natural soil grains of the
Danube river sand.
Lőrin
cz (1986) made a theoretical and experimental study
on the grading entropy. He observed that the
dry density
of the
sand fractions are increasing with grain diameter since the
diameter range of the fractions are doubled with increasing
grain diameter. (This result is paradox if the fractions are
modelled by equal spheres.) He demonstrated this fact by his
minimum dry density
measurements.
The data bases of the foregoing two doctoral programs was
compared first to the Calibration Chamber sand database,
published by Lunne (1992) and Mayne and Kulhawy (1992)
and, a significant difference was found on the databases (Imre et
al, 2011) which was attributed to the dependence on the applied
testing standard for density and to the geological origin.
Therefore, two additional data sets (one composed from
Danube sands and one composed from Bochum sand) were
measured in the University of Bochum. These were statistically
analysed and it was found that the result of the minimum
density measurement of the German Standard DIN 18126 (and
some other standards not using the Proctor mold for the
minimum dry density) may be biased and, as a result, the result
of the minimum dry density test may be dependent on the grain
diameter. The geological origin of the soils was indifferent.
2 MATERIALS AND METHODS
2.1
Variables
The results of the density tests can be expressed in terms of the
void ratio
e
, the dry density
d
, the solid volume ratio
s
or its
inverse, the so called specific volume
v
.
The basic definitions
are as follows:
e
V
V
s
s
1
1 1
(
1
)
s
s
V
V e
s
v
1
(
2
)
where
V
is the total soil volume,
V
s
is the volume of solids and
V
v
is the volume of voids,
d
=
s
s
.
2.2
Density tests
The minimum dry density testing methods are related to a
funnel pouring device, differing in the mold size. The Proctor
mold is used by Kabai (1978)
and Lőrincz (
1986). The diameter
of the mold is 0,625 times the height for the DIN18126, 0,769
times the height for the ASTMD4254 and 0,879 for the Proctor
mold (Fig 1).
Concerning the maximum dry density tests, the similarity
of the Modified Proctor procedure and the Vibrating Table
procedure is kown (Poulos and Hed, 1974). Kabai used an 8
mm thick steel plate was applied on the top of the sample, using
the results of Leussink and Kutzner (1962), to prevent the local
loosening of the sample.
