Abstract:
In geotechnical engineering, the soil type which has had some of the most severe economic
implications is that of swelling clays. Swelling or expansive clays form a subgroup within
the field of unsaturated soils which exhibit large volumetric changes upon wetting and drying,
thereby causing severe distress to structures. The aim of this study was to use a combination of
element testing and centrifuge modelling to investigate certain aspects affecting the behaviour
of piled foundations.
To perform the necessary centrifuge modelling, a sample preparation procedure was developed
whereby the aim was to prepare samples which retained some degree of fissuring. Samples
with such a fabric type facilitate a more rapid ingress of moisture. Element testing was then
carried out to investigate any difference in the mechanical or soil water retention properties of
the laboratory compacted specimens and undisturbed intact samples. The mechanical properties
of primary interest were the pressure required to prevent swell, as well as the clay’s heave
potential at various applied stresses. The results of the element testing conducted illustrated
that the properties of interest remained, for all practical purposes, unchanged between statically
compacted and intact undisturbed samples.
Consolidation testing was also performed on various statically compacted and undisturbed
samples. Additionally, the intrinsic clay properties were established from a test on a reconstituted
specimen and were found to conform well to the framework outlined by Burland (1990).
Comparisons of the consolidation tests on statically compacted and undisturbed samples with
the reconstituted specimen allowed for soil structure and swell induced-softening to be analysed.
This analysis revealed that only undisturbed samples which were inundated at relatively high
stresses existed in structure permitted space upon further loading. This result illustrated the
effect of swelling on soil structure and highlighted how, for the stress range applicable to this study, both undisturbed and statically compacted samples exist in permissible stress states
after swelling. Furthermore, for samples which had undergone swell, yielding was found to
occur at significantly lower stresses than what would be predicted for unstructured soils. More
specifically, yielding occurred well within permissible stress space, rather than on the intrinsic
compression line defined by the reconstituted consolidation test. This finding is consistent with
the swell-induced softening described by the constitutive framework considered in this study.
The centrifuge modelling conducted illustrated that the sample and preparation procedure
utilised, facilitated the ingress of moisture, such that a significant magnitude of swell could
be achieved within a practical time frame. This allowed for various aspects affecting the
performance of piled foundations to be evaluated before (at the clay’s in-situ moisture content)
and after a targeted value of swell was achieved. A series of ‘pull-out’ tests were conducted
to evaluate the shaft capacity of short length piles at various depths in a clay profile, before
and after achieving the targeted swell. The piles used for this evaluation were bored piles
manufactured from rapid hardening grout. From this series of testing it was found that on
average, the shaft capacity of piles reduced after achieving the targeted magnitude of swell.
Exceptions to this trend were found at greater depths in the profile where shaft resistances
increased.
A final centrifuge model was conducted to evaluate the development of lateral pressure
against a pile throughout a swell process. This test included an aluminium pile anchored
at its base, instrumented with lateral load cells. The development of lateral pressures was
monitored up until the targeted magnitude of swell was achieved. This test illustrated a distinct
initial increase in lateral pressure, followed by a more gradual reduction in pressure as swell
progressed. This result helps clarify some discrepancies in the literature on the effect of soil
swell on the shaft capacity of piled foundations. If testing is performed fairly early in the swell
process where lateral pressures are increasing, it is possible that an increase in shaft capacity
will be measured. However, if shaft capacity is assessed at relatively large magnitudes of heave
(as was done in this study), reductions in shaft capacity are likely to be measured.