Environmental Engineering Reference
In-Depth Information
roller, than if the test is carried out on material sampled direct from the borrow area.
The result of this is that the density ratio calculated is lower when the recompacted
soil maximum dry density is used, often resulting in rejection of the fill. Since the
rollers are dealing with the soil from the borrow area, not recompacted soil, this is
unreasonable. There are two solutions for this problem - either lower the density
ratio standard and use the recompacted maximum dry density, or ensure that labora-
tory compactions are done on representative uncompacted material. The latter may
be difficult because of material variability.
(e) Property change on drying. Some soils change properties when dried in an oven or
under lights. Halloysite clays are particularly prone to this, but most clays are
affected. It is desirable not to dry the soil used for the laboratory compaction test com-
pletely, but to the water content needed for testing.
(f) Vibration from nearby construction equipment may affect the density obtained by sand
replacement methods. This can be overcome by using the water balloon method, or by
testing when equipment is not operating nearby.
(g) Inadequate curing of samples. Soils, particularly higher plasticity clays, need time for
water added for laboratory compaction tests to evenly distribute throughout the sample.
This is the reason why most standards require 12 hours to 24 hours “curing” of the soil
before compaction. In a construction situation this may be regarded as impracticable and
not adhered to. As a result, compaction results may be inconsistent and subject to error.
(h) Specification of standard soil tests for “gravelly” materials. The standard tests can be
corrected for the presence of gravel particles up to a reasonable limit. However, if
the soil to be tested is largely gravel size, the potential errors are too great and larger
size compaction moulds must be used, or a methods type specification adopted.
Inexperienced persons may even specify a density ratio rather than density index (rela-
tive density) for granular soils. One needs only to observe the loosening effect of a com-
paction hammer on sand in a compaction mould to appreciate that this cannot work.
14.6.3
Compaction control - some other methods
Some other approaches have been developed for routine compaction control of clay soils,
particularly for fine grained clay soils which are compacted wet of optimum water con-
tent. These include:
-
Specifying water content, and a density ratio based on wet density. The authors have used
this approach for a dam constructed in Papua New Guinea, using halloysitic clay where
previous experience had shown that it was very difficult to define maximum dry density
and water content, i.e. the laboratory testing gave very variable results. The specification
was written to require a water content between 44% and 50% (roughly OWC
3%) for
the mean OWC and the required compaction dry density was 98% of the dry density
achieved in the laboratory at the field water content. This wide range of water content
was only practicable for the soil being used and would not normally be acceptable.
-
Specifying undrained shear strength of the compacted earthfill. Knight (1990) describes
the use of this technique, which is common in dams constructed in the United Kingdom.
Undrained strengths used range from
40 kPa to 110 kPa, with tests being carried out
using triaxial tests on 100 mm diameter driven tube samples, or on remoulded samples.
Lower bound and mean values were specified. Knight (1990) indicates that advantages
of the method are that “measured strengths reflect design intention and testing is
speedy”. The authors' view is that the technique is useful where soils are to be com-
pacted significantly wet of optimum, but that otherwise it is preferable to adopt a den-
sity ratio and water content specification. For most dams, the undrained shear strength
of the core is not a critical issue, because stability is controlled largely by rockfill zones.
