Hydro-Mechanical Properties of an Unsaturated Frictional Material

7.2. SOIL-WATER CHARACTERISTIC CURVE 147
Volumetric water content (%)
Volumetric water content (%)
50
40
30
20
10
0
Scanning imbibition Drainage
θ's= 39%
Loose specimen Initial void ratio = 0.89
0.1 1 10
50
Suction (kPa)
40
30
20
10
0
θ's= 35%
Dense specimen Initial void ratio = 0.66
0.1 1
Suction (kPa)
10
Figure 7.7: Influence of loading path direction (drainage, imbibition process) on the shape of
the soil-water characteristic curve (transient state tests)
Due to occluded air bubbles the initial saturated volumetric water content is not achieved
after subsequent imbibition process. The occurrence of occluded air in the sand column tests I
(transient state test) shows differences ranging from θs = 46% to θ ′ s = 39% for loose specimen
and from θs = 39% to θ ′ s = 35% for dense specimen. However, also the tests performed in
the same testing device (sand column test I) but under steady state condition give similar
results (see Fig. 7.5). The dense specimens show the phenomenon of occluded air to have a
minor effect because the pores of the dense specimens are smaller and therefore the volume
of occluded air appears to be smaller than for the loose specimens.
The phenomena of occluded air bubbles was not found to be significant for the results
observed from the modified pressure plate apparatus (see also Fig. 7.5). The occurrence of
occluded air in the modified pressure plate apparatus tests (steady state tests) has less effect
than in the sand column tests I (transient state and steady state tests). The modified pressure
plate apparatus tests were performed in a small scale apparatus and the sand column tests
I were performed in large scale column. It appears that the size of the equipment influences
the relative effect of occluded air.