Results of property tests on unmodified soils
Soil classification
The results of sieve analysis, natural moisture content, Atterberg limits, and AASHTO
M 145 and USCS classifications of the soil samples are presented in Table 1. Nkporo Shale is classified as Organic Clay under USCS classification and A-7-5(20)
under the AASHTO soil classification system, with the highest Group index value which
represents the poorest subgrade rating. Ekenkpon Shale also classified as OH, while
New Netim Marl classified as MH (Inorganic Silt). Values of maximum dry densities
(MDD) and optimum moisture contents (OMC) obtained from compaction tests are also
presented in Table 1.
Table 1. Soil indices, moisture—density values, and classification for native soil
California Bearing Ratio
Un-soaked and soaked CBR specimens were tested. Table 2 presents the CBR values for native soil samples at optimum moisture content for both
un-soaked and soaked states. These values range from 5 to 26.8Â %. None of the soils
meet basic requirements [AASHTO, and Government of the Federal Republic of Nigeria
(1997]) criteria] or soaked CBR value of 30Â % for both subgrade and subbase.
Table 2. Moisture—density values and California Bearing Ratio for virgin soil
Soil improvement
General considerations
Soil modification can increase the strength properties of a subgrade soil and is usually
carried out using lime, fly ash or cement, or any combination of the three (Little
et al. 2000]), These materials are known as traditional stabilizers (NCHRP 2008]) because they rely on pozzolanic reaction and cation exchange to modify and/or stabilize
soil. Some other materials such as rice husks, sawdust and sawdust ash (SDA) have
also been used, mostly in rural, low traffic volume roads. The choice of improvement
agent is influenced by the following (NCHRP 2008]; Muhunthan and Sariosseiri 2008]);
Soil mineralogy and content (sulfates, organics, etc.…)
Soil classification (gradation and plasticity)
Goals of treatment
Mechanisms of additives
Desired engineering and material properties (strength, modulus, etc.…)
Design life
Environmental conditions (drainage, water table, etc.…)
Engineering economics (cost vs. benefit)
For the present study, all the above were taken into consideration. Ordinary Portland
cement was selected as the improvement agent; lime and fly ash are not readily available
in economical quantities whereas the study area has a number of cement factories.
Strength and resilient modulus requirements for soil stabilization
AUSTROADS research report AP-R434 (Austroads 2013a]) classifies modified or stabilized soil based on water cured 7 and 28-day unconfined
compression strength (UCS) values as presented in Table 3. Generally, there are two broad classifications, namely:
Modified material, with a UCS value less than or equal to 1.0Â MPa at 7Â days or 1.3Â MPa
at 28Â days. The resilient modulus for the soil should be ?1500Â MPa (1.5Â GPA).
Bound material, which is further subdivided into;
1. Lightly bound, with 7-day UCS between 1 and 2.0Â MPa or 28-day UCS between 1.3 and
3.0Â MPa with thickness less than 250Â mm. The resilient modulus value should range
from 1500 to 2000 MPa (1.5–2.0 GPA).
2. Heavily bound with 7-day UCS value 2.0Â MPa or above 3.0Â MPa at 28Â days with a
thickness 250 mm and design deformation modulus between 2000 and 20,000 MPa (2.0–20.0 GPa)
Table 3. Typical properties of modified, lightly-bound and heavily-bound materials
Based on the above and the UCS values listed in Table 4; at optimum cement content (the most economical amount of cement that gives a resulting
mix a minimum unconfined compressive strength value of 1.5Â MPa), Nkporo shale has
a maximum UCS value above 3.4Â MPa at 12Â % cement content, New Netim Marl 2.17Â MPa
at 3Â % cement content, and Ekenkpon 1.81Â MPa at 7Â % cement content. Therefore Nkporo
shale can be classified as heavily bound, while New Netim Marl and Ekenkpon shale
soil can be said to be lightly bound. On the other hand the NCHRP (2008]) only distinguished between modified soil and stabilized soil based on UCS value
of 1.5Â MPa. Based on this criterion all the soils at the optimum cement content are
stabilized soils. The AUSTROADS research report, AP-R434 (2013a]) classification appears to be superior over the NCHRP (2008]), since it further classifies soil into lightly bound and heavily bound based on
the UCS, and also recognizes that bound materials contribute significantly to pavement
strength by stiffening the pavement foundation. However this stiffening can be accompanied
by development of shrinkage cracking. NCHRP (2008]) also recognized this and suggested that slow setting additives should be added to
such stabilized soil.
Table 4. Unconfined compressive strength and Young’s modulus values for the three soils at
different cement contents
Strength and durability test results
Table 4 presents the minimum and peak values of unconfined compression tests results for
different percentages of cement for each of the soil. Figure 4 shows the variation of unconfined compressive peak strength values for the three
soils with various percentages of cement. ASTM D 560 (2003]), and Austroads research report AP-R434 (2013a]), specify unconfined compressive strength (UCS) value of 1.5Â MPa on 28Â days moist
cured sample. A durability line based on the above requirements is shown in the Fig. 4. The different values of UCS for each soil were plotted against the different percentages
of cement. Best-fit straight lines were determined for each soil using least-squares
regression. The r
2
correlation coefficients were 0.84 for New Netim Marl, 0.82 for Ekenkpon shale, and
0.31 for Nkporo shale. The minimum cement content for each soil that will ensure durability
of the stabilized soil in the field is found where the correlation line crosses the
durability line. These values are, 2.20Â % for New Netim Marl and 6.2Â % for Ekenkpon
shale. The resulting soil mixtures in both cases at the optimum cement content have
low plasticity values; tending towards 7Â % in case of New Netim Marl and 9Â % in case
of Ekenkpon shale. Such value could not be put forward for Nkporo shale since the
single linear correlation coefficient is poor. Considering the data for Nkporo shale
as two sets; the first, from 0 to 7Â %, and a second data set for 10 and 12Â % cement
content. A bilinear correlation for the two sets resulted in a second correlation
line that gives optimum cement around 11Â %. The maximum value of UCS test for the
10Â % is 1.5Â MPa, whereas the 3.41Â MPa, for 12Â % cement content surpassed this minimum
value. Also no experiment was carried out for 11Â % cement content; therefore 12Â %
by weight of cement content will serve as the optimum cement content required to stabilize
Nkporo shale.
Fig. 4. Variation of unconfined compressive strength with cement content for the soils
Soil indices and stabilization
Table 5 presents variations of Atterberg limits for the three soils with various percentages
of cement content. The tests for determination of Atterberg limits were carried out
on the 28Â day water cured sample. Compared to its virgin sample the Nkporo shale plastic
limits fluctuates in values as cement content increases. LL varies with increased
cement content from 52Â % at zero cement to 37Â % at 12Â % cement. Plasticity Index (PI)
reduces from 22Â % at zero cement to 9.01Â % at 5Â % cement, jumps up to 21.0Â % at 10Â %
cement content, then 11.5Â % at 12Â % cement content. The increase in plasticity Index
at 10Â % cement content for the Nkporo shale is likely due to an optimum reaction with
sulphate present in the shale (Nganje et al. 2015]). Sulphate levels at concentration of 3000Â ppm (3000Â mg/l) and above (NHRCP 2008)
can react with cement to produce a swelling potential that is destructive to pavement.
The various degrees of concentration of sulphate measured in ground water (hand dug
wells), and surface waters (ponds, streams) within Nkporo shale geologic area reported
by Nganje et al. (2015]), ranges from 0.00 to 8542.80Â mg/l. The value that must be dominant in the Nkporo
soil in the area must be less than the threshold value for which the soil reaction
with Portland cement will result in destructive expansive potential but significant
to raise the PI of the stabilized soil at 10Â % cement content, which is reduced when
cement reaches 12Â %. This trend is similar to the pattern shown by unconfined compressive
strength test, the result of which drop to a minimum with progressive increase in
cement content that reaches the lowest at 10Â % cement content and then increasing
significantly at 12Â %. An explanation for this behavior trend will be made in the
discussions under ‘cement stabilization and pH’.
Table 5. Variation of Atterberg limits and plasticity index for the three soils with different
percentages of cement content
The New Netim Marl soil shows progressive decrease in PI from zero cement content
up to 5Â % cement content by weight of the soil. With values of 15Â % at zero cement
content to 11Â % at 2Â % cement content then to 7Â % at 3Â %,and 5Â % of cement content.
The value then goes up to 17Â % at 7Â % cement content indicating a less stiff soil.
Analysis of strain existing at the peak stress shows that at 5Â % cement content the
peak stress is 2498.33 kPa and the associated strain is 8.80 × 10
?03
and at 7Â % cement content the peak stress is 2822.9Â kPa and the associated strain
is 1.29 × 10
?02
. This strain value is larger than the one at 5Â % cement content. The Atterberg limit
values for this stabilized soil vary in a fluctuating manner, though in a cyclic pattern,
from 48Â % at zero cement content to a lowest value of 37Â % at 5Â % cement content and
then up again to 47Â % at 7Â %.
The Ekenkpon shale soil shows an initial increase in LL at 2Â % cement, but further
increase in cement content results in continual decrease in LL until 7Â % when the
liquid limit attained was 40.6Â %. The PI follows a similar trend; with a PI of 28Â %
at 2Â % cements content decreasing to 9Â % at 7Â % cement content.
Cement stabilization and pH values
Portland cement is comprised of calcium-silicates and calcium-aluminates that hydrate
to form cementious products. Cement hydration is relatively fast and causes immediate
strength gain in stabilized layers (Little et al. 2000]). The mechanism of Portland cement reaction with soil is well documented. When Ordinary
Portland cement is mixed with soil in the presence of adequate moisture, cement hydrates
and produces free lime, Ca(OH), which reacts Pozzolanically with soil as long as the
environment is alkaline, that is with a high pH value. It therefore follows that a
soil with pH value in or towards alkaline region will react more productively in the
presence of adequate moisture. Hence the need to determine the pH values of the native
soils being investigated.
Soil pH values were; 2.27 for Nkporo shale, 6.7 for New Netim Marl, and 4.28 for Ekenkpon
Shale. Typically, a soil with low pH (acidic) will require more cement content for
stabilization as its pH will have to be raised to significantly more than ‘7’ before
cement will react properly (the pH of cement typically ranges between 9 and 11). Conversely
a soil with a pH value towards or in the alkaline region may require less amount of
cement than the previous one. The Nkporo shale with a pH of 2.27 did not show any
appreciable strength gain with the initial percentages of cement (2, 3, 5, 7Â %) used
for all the soils. Hence the cement content utilized for the Nkporo shale was increased
to 10 and 12Â %; whereas New Netim marl with a pH value of 6.7 required only 3Â % cement
content by weight for stabilization, and Ekenkpon shale with a pH of 4.25 required
7Â % by weight of cement. Although the clay mineralogy of a soil also influence the
way it will react with cement; the mineral structure of the soils under study were
not determined as it was not one of the objectives of this work. Therefore no comment
on the influence of clay mineralogy on their stabilization can be made.
California Bearing Ratio and cement content
Although the CBR value is not the main criteria used in determining the optimum cement
content for stabilization; it was determined for some trial mixes that were soaked
for 28Â days. Ekenkpon Shale which occurs more extensively in the study area and with
a pH value in between the other two was mixed with the trial percentages of cement.
Table 6 presents CBR results for these trial mixes. Figure 4 shows a correlation graph between CBR and percentages of cement. The 28 day water
cured soaked CBR value increased with increased cement content from 10.5Â % at 2Â %
cement to 25.8Â % at 4Â % cement to 49Â % at 5Â % cement.
Table 6. Variation of California Bearing Ratio values with cement content for Ekenkpon shale
soil (28Â days soaked)