Environmental Engineering Reference
In-Depth Information
Figure 15.10.
Simplified method of calculating rockfill modulus during construction, and pseudo mod-
ulus during reservoir filling (Fitzpatrick et al., 1985, reproduced with permission of
ASCE).
Equation (15.1) does not allow for the distribution of vertical stress in the dam due to
the shape of the dam. E
rf
is not a true modulus of the rockfill, and should only be used as
discussed below to estimate the face slab deformation.
The following points can be made:
- The use of high strength rock does not guarantee a high modulus, e.g. Foz do Areia
Dam. The relatively low modulus for this dam is due to poor grading, i.e. a lack of sand
size particles (Pinto et al., 1985), and is typical of basaltic rocks in that region. The void
ratio of the compacted rockfill is a guide to the rockfill behaviour (note that Foz do
Areia Dam has a relatively high void ratio of 0.33. Most rockfills have a void ratio less
than 0.25).
- Low strength rocks, which break down significantly, e.g. Kangaroo Creek, Little Para
and Mangrove Creek, can give quite high moduli, but it must be remembered that they
will probably also give low permeability.
- The highest moduli are achieved for gravels, where the rounded shape limits crushing
of the contact points between particles of the fill. Conversely the high stresses on the
contact points in the poorly graded basaltic fill help cause the low modulus.
-Well graded high strength rockfill can also give high moduli, e.g. Murchison, Bastyan and
Reece.
- Additional compaction (e.g. Murchison 8 passes) and thin layers (e.g. Alto Anchicaya)
can lead to higher moduli.
- The modulus is dependent on layer thickness, e.g. at Foz do Areia and Khao Laem dams
where Zone 3B, compacted in layers twice the thickness of Zone 3A had moduli 60% to
80% of that for Zone 3A.
-
The pseudo modulus on first filling is commonly 2 to 3 times that observed during con-
struction. This is important as the face slab displacement is inversely proportional to
the “first filling” modulus and is discussed further in Section 15.2.5. Hunter (2003),
and Hunter and Fell (2002) studied this and concluded that the apparent increase in
modulus is an artifact of the method of calculation, rather than real. They showed that
the layering effects within each rockfill layer had little effect and that, while the mean
stresses increase from the construction condition as the water load is applied, the
change in deviator stress is small.
It will be noted that in all dams water was added to the fill, usually at a rate of 15% to
25% of the rockfill volume. This results in reduced compressibility, although the
improvement may be marginal in some cases.
Cooke and Sherard (1987) conclude that: “(1) For most hard rocks and CFRD of low
to moderate height, the addition of water has negligible influence on the dam behaviour.
(2) For high dams and for rocks which have significantly lower unconfined compression
