Environmental Engineering Reference
In-Depth Information
- Only a limited number of cases of well-compacted gravels were available; insufficient to
undertake analysis. These generally have significantly greater secant moduli at any given
stress level in comparison to quarried, very high strength rockfill. Where the gravel is of
finer size (e.g. Crotty Zone 3A) the use of a large scale laboratory testing that virtually
encompasses the field particle size distribution would provide suitable estimates of mod-
uli, but only if the laying and density in the field are reflected in the laboratory testing.
Alternatively plate load tests on gravels compacted in field trials could be used;
- Monitoring should be undertaken during the embankment construction as a check on
the pre-construction estimation of deformation.
Data to allow calculation of the moduli of dumped rockfill during construction or first
filling was not available to Hunter (2003). It could be anticipated that the modulus would
be much lower than for compacted rockfills. Dumped rockfill has a much greater suscep-
tibility to collapse on saturation (e.g. Cogswell Dam (Bauman, 1958) settled 5.4% fol-
lowing a heavy rainstorm which saturated the rockfill which had been dumped without
sluicing). Back-analysis of the deformation of a narrow central core earth and rockfill
dam built of dumped, sluiced, somewhat weathered basalt rockfill implied an equivalent
secant modulus of 5 MPa for the rockfill on the upstream side, but probably higher on the
downstream side. Rockfill at El Infiernello dam which was dry dumped and spread in 2 m
lifts gave a modulus of 17-27 MPa (average 22 MPa). At Ita dam, rockfill dumped into
10 m of water gave an estimated modulus of around 15 MPa to 19 Mpa.
For assessing the effects of embankment raising it is recommended that moduli for the
rock-fill be determined from the deformation performance during construction, if that is
available. Alternatively, tangent modulus versus applied vertical stress can be obtained
from integration of this stress-strain curve. The variation can then be modelled by con-
sidering the embankment in a series of layers of varying moduli. For a more rigorous
analysis numerical modelling can be used.
15.2.4.2
Estimation of the first filling “pseudo modulus” E
rf
Hunter (2003), Hunter and Fell (2002, 2003c) showed that there was a relationship between
the ratio of the pseudo modulus on first filling (E
rf
) to the representative modulus during con-
struction (E
rc
) with embankment height and upstream slope as shown in Figure 15.12.
4.0
Khao Laem
Estimation of trendline for upstream
slopes of 1.3-1.4H to 1V
3.5
3.0
Scotts Peak
2.5
2.0
Ita
Mackintosh
Foz do Areia
Golillas
1.5
1.3H to 1V
1.4H to 1V
1.5H to 1V
1.6H to 1V
Crotty
1.0
Serpentine
Very approximate estimation of trendline for
upstream slopes of 1.5H to 1V
0.5
0.0
0
25
50
75
100
125
150
175
200
Embankment Height (m )
Figure 15.12.
Ratio of pseudo modulus on first filling and construction rockfill modulus (E
rf
/E
rc
) versus
embankment height. (Hunter 2003; Hunter and Fell, 2002, 2003c).
