43
pressure/volume controllers (PVC) and the soil suction measuring device. Schematic
diagram of the RLT test apparatus is shown in Figure 28.
Figure 28.
Schematic diagram of the suction-controlled RLT test apparatus.
For both of the subgrade soils, the resilient modulus tests
were conducted at a wide
range of suction (moisture content). From the RLT tests data, it was observed that the
resilient modulus of subgrades were highly dependent on the net bulk stress, deviator
stress and the matric suction (Figure 29).
Figure 29.
Resilient modulus of the Luleå subgrade soil at different suction states
(degrees of saturation).
44
A literature survey was conducted and several recent suction-incorporated
resilient
modulus models were summarized (Parreira and Goncalves, 2000; Khoury and Zaman,
2004; Yang et al., 2005; Liang et al., 2008; Cary and Zapata, 2011; Nowamooz et al.,
2011; Ng et al., 2013).
The dataset from the suction-controlled resilient tests were then used to calibrate
several recent predictive models that account for matric suction of the soil and evaluate
the capability of them in capturing the resilient response of the material under different
load and moisture conditions.
Main findings in Paper III and IV
The resilient modulus of the subgrades increased with increasing net bulk stress
and decreased with increasing deviator stress (octahedral shear stress).
Resilient modulus increased significantly with increasing matric suction.
Change in the resilient modulus with varying net bulk and deviator stress was
generally more significant at higher suction values (the degree of nonlinearity of
the material stress-strain relationship was more significant at higher suction
levels).
The resilient modulus predictive models that combined the three fundamental
stress state variables (net confining pressure, deviator stress and matric suction)
performed more satisfactorily in capturing the resilient
response of silty sand
subgrade soils and their variations due to seasonal changes in the moisture
content.
Given the goodness of fit of the suction-resilient
predictive models, it can be
concluded that considering matric suction as an independent
stress state variable
in the models might be a rational approach for incorporating moisture effects in
the predictive models.
Dostları ilə paylaş: