Lateritic (LAT) soils are formed from a chemical weathering process in the tropics and
subtropics, which usually results in leaching out of silica and the relative increasing
concentration of iron and aluminium sesquioxides. As a result, there is the formation of iron
nodules or cemented aggregates, owing to the different polarity of the surface charges of clay
and sesquioxides. The particle aggregations are usually weakly cemented and are known to
break down into finer particles due to thermal and mechanical working. This creates difficulties
with defining a particle size distribution and soil classification, thereby making it difficult to
predict their engineering behaviour. Moreover, LAT is naturally found unsaturated, thus
suction would play a major role in their hydro-mechanical properties under working conditions.
Understanding the dynamic properties and small strain behaviour of LAT soils is essential for
assessing their working performance since it is widely used as fill materials for subgrade
construction of highway and railway, foundations and retaining walls.
The effects of suction on the dynamic properties and small strain behaviour of unsaturated LAT
soils are rarely studied. In this study, resonant column tests were carried to examine the
influence of suction on the dynamic properties of the two compacted lateritic sandy clays from
Nigeria (NL) and Ghana (GL). In addition, a series of suction-controlled constant p shearing
test was carried out to examine the influence of suction on small strain shear behaviour of the
compacted NL. The small strain behaviour of NL is compared with that of another soil, which
was tested following a similar stress path in a previous study. Moreover, the microstructure of
the soils was studied through scanning electron microscope (SEM) images.
From the resonant column tests, it is found that the initial shear modulus (G
0) of both NL and
GL increases with increasing confining pressure and suction. However, the influence of suction
on the G
0 values is larger for GL than NL due to the existence of many smaller aggregation of
GL. An increase in suction to 300 kPa for both soils resulted in a lower elastic threshold shear
strain and a shift of the normalised shear modulus (G/G
0) degradation curve towards lower
shear strain values different from other soils. This result is most likely due to microstructural
evolution due to drying to 300 kPa suction. For the damping ratio (D), it was found that the
initial damping ratio (D
0) for NL increase from 10% to 23% whereas for GL, the D
0 values
increases from 11% to 14% after drying to 300 kPa. The difference in D
0 for NL and GL with
suction is attributed to larger aggregation of NL because of its higher iron sesquioxide content
causing more cladding. Thus, likely more shrinkage and an enhancement in larger aggregation
during drying resulting in a significant increase in D
0 of NL.
Results from suction-controlled constant p shearing showed that the secant shear modulus
(G
sec) of NL increases by about 50% when suction increases from 1 to 150 at a shear strain of
0.003%. With a further suction increment of 150 kPa, G
sec increases by about 10% only.
Moreover, NL is less dilative at higher suction, different from a complete decomposed tuff
(CDT) in a previous study which was observed to be more dilative as suction increases soils.
This behaviour likely because NL is highly aggregated, as revealed in the SEM images. When
suction increases, water between aggregates loses quickly, even though intra-aggregate pores
are likely to hold water. As a result, suction effects on the NL shear modulus are therefore more
significant at lower suctions (1-150 kPa) but negligible at higher suctions (150-300 kPa).
The measured results in this study suggests suction and microstructural evolution dependency
for LAT soils. Hence, to describe the small strain behaviour of unsaturated LAT soils, structural
parameters which considers microstructure evolution with suction on shear modulus, shear
modulus degradation and damping characteristics should be considered.
Post a Comment