275
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
1
Laboratory testing issues related to crushable sands
Questions concernant des essais de laboratoire sur les sables écrasables
Wils L., Van Impe W.F., Haegeman W.
Ghent University
Van Impe P.O.
AGE (Geotechnical and geo-environmental consultancy)
ABSTRACT: Working with crushable sands, laboratory test issues arise related to their crushable nature and specific morphology.
Complications occur already when determining the minimum and maximum density. The major issue is erosion of the sand during
transportation and flow at testing, resulting in an ever increasing maximum density, affecting the definition of relative density,
D
r
. In
addition, as with any other granular material at the initial stages of compression, a crushable sand densifies when interparticle voids
decrease; yet when at a relatively low stress level particles also start to crush, their so-called intraparticle voids – typical for sands
from bioclastic origin – gain importance. Other issues influencing the crushable sand matrix are: polishing of angular grains into more
rounded particles; larger shells sheltering smaller grains; apparent cohesion due to interlocking of the angular particles, creating
sandclusters with the appearance of larger particles; etc. During any kind of treatment of the crushable sand samples, attrition of the
particles must be addressed with great care and granulometric properties should be closely monitored throughout.
RÉSUMÉ : Pendant les essais de laboratoire sur le sable écrasable, beaucoup de questions se posent liées à sa nature déformable et sa
morphologie spécifique. Des complications surviennent déjà lors de la détermination de la densité minimale et maximale. Le
problème majeur est l'érosion du sable pendant le transport, ce qui augmente continuellement la densité maximale, et influence la
densité relative,
D
r
. En outre, comme dans autre sols granulaires dans les premières étapes de compression, un sable déformable
densifie par diminution des vides interparticulaires; pourtant, à partir d'un niveau de contrainte relativement bas les particules
commencent aussi à s’écraser, puis les vides dits intraparticulaires – trait des sables bioclastiques – rendent important. D'autres
questions aussi influent la matrice des grains écrasables : le polissage des grains anguleux; grandes coquilles abritant des petits grains
de sable; la cohésion apparente par l’emboîtement des particules angulaires, ainsi créant des clusters qui ressemblent à de grosses
particules; etc. Pendant tout type de traitement des sables écrasables, l'attrition des particules doit être abordée avec soin et les
propriétés granulométriques doivent être surveillés.
KEYWORDS: Crushable sands ; Relative density ; Sieving ; Breakage.
1 INTRODUCTION
At the Ghent University Laboratory of Geotechnics, an on-
going research program focuses on the stress-strain behaviour of
crushable sands. Several laboratory test issues inherent to the
brittle nature and specific morphology of these grains
complicate the performing and interpretation of experiments
and the deduction of soil mechanical properties. Tests on
crushable sands need an appropriate approach that differs from
non-crushable, silica sands.
2 PREPARATION OF SAMPLES FOR LABORATORY
TESTING
Since tests in soil mechanics laboratories represent geotechnical
situations in the field, samples need to be prepared at a relevant
density. Usually the minimum and maximum dry density of a
sand are determined according to standard methods, and a
sample is prepared at the density of interest, and possibly
saturated if it is to resemble offshore conditions.
The minimum density in the research on crushable sands is
determined following the ASTM D4254. Method A, herein
described consists of pouring a mass of sand into a mould with
known volume through a funnel. The pouring spout of the
funnel has an inside diameter of either 12.7 mm or 25.4 mm,
depending on the grain size of the sand. Should the grains be
small enough, the small funnel can be used and this will yield
the lowest density owing to blocking of the sand and a very
slight sand stream. However, for a fine calcareous sand the
small funnel gets blocked due to a cohesive arch formed by the
angular particles. Jamming can only be avoided by gradually
filli
ng the funnel, along with the progression of the sand
through the outlet. Therefore, contrary to what is normally the
case, lower densities are obtained using the large funnel and
pouring sprout.
For determining the maximum density of a crushable sand, the
usual Proctor compaction is replaced by a less invasive
vibratory table densification process as described in ASTM
D4253. In this procedure a known mass and volume of sand is
compacted by fixing the mould onto a table that vibrates at a
frequency of 50 Hz for 12 minutes. A surcharge of 13.8 kPa
keeps the sand grains from segregating. The test is repeated at
different peak-to-peak amplitudes between 0.30 and 0.91 mm,
where
the optimal amplitude is the one where the
energy transfer of vibration to compaction is most efficient,
providing the maximum density by definition. This procedure
only holds true as long as there is no crushing during vibration.
Occurrence of crushing was evaluated by comparison of grain
size distributions before and after the densification. Instead of
having to divide the whole
3.5 kg of sand into manageable
portions to sieve, it was originally assumed that crushing would
occur mostly in the top layer; hence a 120 g sample was taken
from there. The granulometry then showed no evidence of
crushing and the test could therefore be continued at higher
