1063
Technical Committee 106 /
Comité technique 106
2 HYDRAULIC BEHAVIOUR, RETENTION
PROPERTIES AND FLOW
Nearly one third of the papers submitted to the Discussion
Session deal with retention and flow properties, hence
recognising the role of hydraulic state variables on the coupled
hydro-mechanical response of unsaturated soil systems.
Characterisation of the retention properties is tackled at
different scales.
Carrillo-Gil & Carrillo-Acevedo
(Peru)
summarise 20-years data and models at regional scale for
tropical Peruvian soils in the Amazon region (Fig. 1). Both
correlations based on soil index properties and experimental
data from suction cells are analysed, to provide a general view
of the retention properties of three classes of soils (clayey –
silty– sandy) coming from five different regions.
When the regional scale is analysed, only results for water
content are given, irrespective of hysteresis of the soil water
retention mechanisms, of the void ratio and of the hydraulic
path, hence disregarding the coupled evolution of volumetric
strain and water content. The estimated water retention curves
(WRC) typically give wide ranges of water content for given
suction, although typical patterns can be identified for the
different soil classes (Fig. 2). As it is often the case, silty soils
are the most difficult to be uniquely characterised, due to wider
differences in plasticity, void ratios and fabric of the silty soils.
Differences come from both heterogeneity of the soil properties
and different initial void ratios. Attempts to preliminary
characterisation of retention properties at the regional scale may
be of relevant use for risk mapping. Hopefully, similar
databases should be enriched in the future, and possibly re-
analysed by means of statistical tools, to fully exploit their
potentialities.
.
Figure 1. Regional scale for hydraulic characterisation of tropical
Peruvian soils (
Carrillo-Gil
&
Carrillo-Acevedo
)
.
Figure 2. Proposed WRC ranges for the Huallaga River Watershed:
A-sands; B-clays; C-silty-clays (from
Carrillo-Gil
&
Carrillo-Acevedo
)
Reducing the scale of investigation,
Sugii et al.
(Japan)
analyse a mixed experimental and numerical approach to study
the retention and the conductivity properties of an upper
unsaturated sandy layer at the site scale, with an infiltrometer
scheme typically coming from the field of hydrology. The
Authors suggest that simple approximations for the pressure
field and moisture distribution upon infiltration may be
sufficient to get reasonable estimates for the unknown variables,
provided suction is measured at a convenient depth. The
comparison between the field data and the results of a
laboratory model, replicating the experimental procedure in situ,
shows that while the two hydraulic conductivity functions
compare well, WRC estimates present significant differences,
possibly coming from air entrapment effects.
The latter observation suggests that characterising the
hydraulic properties of unsaturated soils still presents open
issues, under different viewpoints. On the one hand, faster
experimental procedures are sought, in order to allow for
reasonable costs – especially in terms of time – of experimental
tests. On the other hand, proper characterisation of the hydraulic
properties of unsaturated soils need correct interpretation of the
multiphase flow process promoted by the different experimental
procedures.
For most soils, the determination of retention properties on
the whole range of possible suctions usually requires a
combination of different experimental techniques, possibly
controlling different exchange mechanisms.
Maček et al
.
(Slovenia) combine data from two commercial equipment, to
investigate the drying branch of the WRC of different soils. The
WRC data are derived from tensiometer readings in an
evaporation device for low suctions (0÷0.2 MPa), and from
relative humidity in the high suction range (>1 MPa), by means
of a dew-point potentiometer. By extrapolating the calibration
range of the evaporation apparatus, they show how the two data
sets may provide a reasonable picture of the whole drying
branch of the WRC.
Reis et al
. (Brazil) discuss how the experimental
determination of the drying branch of the WRC can be speeded
up by imposing suction with a small commercial centrifuge,
carefully enhanced for WRC testing. Although the original idea
dates back more than one century, still the procedure is not
common in unsaturated soil testing, and it deserves further
attention. In the equipment described by the Authors, the
equivalent suction can be controlled either by changing the
angular velocity of the centrifuge, or by increasing the lever arm
of at constant rotational speed. Four samples could be tested at
the same time, after careful assembly and saturation of the set-
up. Results obtained in the small centrifuge compare well with
data from more traditional testing procedures on both
undisturbed and remoulded samples of clayey silty sand in the
range 0÷0.9 MPa, even though small differences between
results of the various techniques may be observed, especially in
the low suction range. The evidence suggests that careful
inspection of the influence of volume changes on the state
variables during testing is necessary to interpret correctly data
from different testing methodologies. Also, data obtained in the
laboratory are seldom the result of the behaviour of the soil
samples alone, but they are affected by the whole experimental
set-up, materials and procedures adopted. The work presented
by
Nishimura
(Japan), in which a micro-porous membrane is
tested in a pressure plate apparatus on different soil types,
confirms the latter observation.
Innovative and promising techniques for unsaturated soils
now try to exploit electromagnetic properties of multiphase
mixtures.
Toll et al
. (UK) discuss an efficient fast multi-
electrode resistivity system, and present a new combined sensor
for suction and water content (Fig. 3), consisting in a coiled
TDR, which can be used in conjunction with a high capacity
tensiometer for simultaneous measurement of water content and
pore water pressure.
