Environmental Engineering Reference
In-Depth Information
of intergranular porosity of sands. Pressure solution effects should proceed much more
rapidly in carbonate sands than in quartz sands, because the solution rates of the carbon-
ate minerals are much higher than that of quartz. Relatively inexpensive research (e.g.
examination of thin sections cut through epoxy-impregnated old pavements) may show
whether or not the effects are rapid enough to render filters composed of
Category O
car-
bonates ineffective.
3.7.7.2
Category Y carbonate materials
(a)
Change of grading due to dissolution
. Sedimentary geologists consider that in fresh
water high magnesian calcite is more soluble than calcite and aragonite (Prothero and
Schwab 1996). Quantitative data on the solubility and solution rate of high magne-
sian calcite are not presently available. However, from the field and laboratory
evidence discussed below, the authors believe that
Category Y
carbonates used as fil-
ter materials would suffer more rapid dissolution than equivalent
Category O
car-
bonates.
(b)
Partial dissolution and recementation
. Netterberg (1975) discussed very high CBRs
yielded by pedogenic calcretes in Southern Africa and stated
“It is difficult to ascribe
soaked CBRs much above those yielded by crusher-runs to anything other than
cementation”
. He also included test results showing that the CBRs of some calcretes
had more than doubled after a few cycles of wetting and drying. In Netterberg (1969)
he described studies aimed at understanding how these calcretes were formed in soil
profiles. He concluded that fine carbonate material in the soil is dissolved and rede-
posited as cement during changes in soil water suction and partial pressure of carbon
dioxide. He also provided estimated rates of calcrete formation in nature, ranging
from 0.02 mm/year to 1 mm/year (Netterberg 1969, 1978). In Netterberg (1971) he
suggested that the apparent “self stabilisation” of calcretes used in pavements may be
caused by similar (solution and redeposition) processes. The authors have seen many
examples of recemented calcrete pavements in Australia.
Sterns (1944) and Tomlinson (1957) describe evidence of recementation of coral
sands and gravels, where used as base materials in airfield pavements.
Coquina limestone of Tertiary age is used widely (when crushed) as base course for
highways in Florida. This rock comprises corals, shell fragments and quartz grains.
Graves et al. (1988) report that the strength increase which occurs with time in these
pavements is due to dissolution and reprecipitation of fine carbonate particles. They
support this claim using results of the LBR (limerock bearing ratio) test, a modified
version of the CBR. Quartz sands were mixed in different proportions with carbonate
sands containing less than 2% of fines. The samples were compacted into LBR moulds
and tested after soaking continuously in water for periods of 2, 7, 14, 30 and 60 days.
All samples with 40% or more of carbonate sand showed strength increases ranging
from 23% to 65%, after the first 14 days of soaking. No further increase was recorded
after this time. Graves et al. suggested that the very fine carbonate particles may have
been “used up” by the end of 14 days and that the coarser grains were not providing
cementing material as efficiently. Tests with crushed carbonate base course materials
containing “slightly more carbonate fines” were tested similarly for 30 days and
showed increases in strength up to that time.
McClellan et al. (2001) carried out further laboratory research into these effects but
were unable to reproduce comparable strengthening of samples within curing times of
up to 60 days.
(b)
Interlocking of grains due to pressure solution.
Shoucair (2003) has advised that the
Florida Department of Transport has been adding coral sands to their road bases with
significant strength gains. However he notes that the gains “practically disappear
when the materials are soaked again” and considers that mechanisms described by
