Abstract:
Expansive soils are recognized as problematic soils that impose several challenges for civil
engineers. Such soils undergo significant volume change in case water penetrates into them, and
they shrink as they lose moisture. Lightly-loaded engineering structures such as pavements,
single story buildings, railways and walkways may experience severe damages when they are
founded on such soils. Therefore it is necessary to develop models to predict swelling pressure
and swelling potential of expansive soils.
This research presents an experimental investigation of swelling behavior (swelling pressure and
swelling potential) of expansive soils. The expansive soil specimens were prepared in the
laboratory. Atterberg limits, Grain size distribution, Activity, Optimum water content, Maximum
Dry density, Swelling pressure and swell potential were determined.
The correlations between the swelling behavior of test samples and fundamental properties of
test samples were studied. Additionally, the correlations between swelling behavior and some of
the index properties of test samples were investigated.
The model was formed in one equation with many parameters. The main advantage of this model
is that only one parameter is needed to run the model. The main parameters that are needed to
run the model are the applied pressure and initial dry density. The other parameters of the model
including the initial slope of the swell-time curve, the final slope, the reference swell, and the
peak swell were investigated experimentally.
Based on the experimental results, empirical relationships were developed for determination of
these parameters as functions of the applied pressure and initial dry density. The model
prediction were compared with the experimental results and showed good agreements for all
levels of applied pressure and initial dry density of the tested soil.
As final conclusion of this research, the values of swell pressure and swell potential of the test
samples from the experimental investigation are compared with the predictive values of the same
based on currently proposed and other suggested models.
