Introduction

The highway that links the states of Abia and Akwa-Ibom in southeastern part of Nigeria is at present a two-lane, two-way carriage way. The road is presently in a bad state with many failures in more than eighty percent of its total length of 34 km. The road has significant economic importance due to the fact that it also serves as the only gateway from the southern end of the country into Cross River state, the state bounding Akwa-Ibom at its Northern end. The highway, therefore, is critical to movement of goods and persons between the three states and to the remaining part of the country. The first 16 km. of the highway is presently being reconstructed into a dual carriage way. This paper presents the geotechnical characterization of alignment proposed for the remaining 18 km.

Objectives

The objectives of the geotechnical characterization are:

  1. 1.

    Identify soil domains in the proposed alignment and their extent in width and depth along the alignment.

  2. 2.

    Determine engineering properties of the soil domains based both on field and laboratory tests.

  3. 3.

    The influence of the engineering properties of the soil domains to the proposed construction and eventual behavior of the highway pavement.

Site description and geology of the study area

The proposed dualisation alignment follows the existing road in most part. The existing alignment is crossed by two seasonal streams with associated bridges. The geotechnical characterization at these two bridge sites crossing the two rivers are excluded from this investigation. Generally, the alignment has flat topography with no undulations except at the seasonal stream sites where approaches are characterized with valleys. The geographical co-ordinates of the alignment termini are 05° 07′ 9.9″ N, 07′ 22" 17.52″ E for Aba, and 05° 08′ 28.9″ N, 07° 33′ 34.99″ E for Ikot-Umo Essien.

According to the Nigerian Geological Survey Agency 2006 base map of Akwa–Ibom State, the geology of the area spans from Cretaceous through Tertiary to Quaternary. In the Cretaceous, the following in younger succession characterize the area. These are the Asu River Group, Eze Aku Shale Group, Lower Coal Measures, False Bedded Sandstones and Upper Coal Measures. In the Tertiary also, in younger succession, the Imo Shale Group, Bende Ameki Group, Coastal Plain Sands, and Alluvium constitute the Formations within the study area. The surficial geology along the alignment in general is the clastic sediments of Coastal Plain Sands and the Lignite Series Formations which have ages in the range between Tertiary to early Quaternary. Lithology essentially is sands and clay in the former, and lignite, claystone and shale in the latter Formation. These two formations form the youngest stratigraphic units in the area.

Penetrometers

Different types of penetrometers are in use for foundation investigation and characterization, the most common being the Dutch Cone Penetrometer (CPT) and Standard Penetrometer (SPT- ASTM D-1586 2011). Advances in geotechnical instrumentation have seen many improvements in both penetrometers, but more have been made on the Dutch Cone Penetrometer (CPT) than the SPT. The CPT has seen improvements and developments that increase its capability to handle differing ground conditions (Campanella et al. 1982), and accurately identify soil types (Begemann 1965, Olsen and Farr 1986; Robertson 1990), and give better estimates of relevant geotechnical soil parameters. For geotechnical characterization of a highway, the SPT is a bulky equipment to deploy. Although depth of investigation could be 30 m or more, which is far higher than what is usually required for a highway pavement (FHWA NHI–05–037 2006), The depth of investigation is based on zone of influence, which for the asphalt pavement is about 3.81 m (Vandre et al. 1998). The result obtained with SPT is usually bearing stress or bearing capacity. The same is true for the CPT, although it is easier to deploy the CPT, than the SPT. Both of them are capable of deep investigation which comes into play in that part of a highway that needs embankment fillings, or at potential bridge sites. For highway geotechnical investigations other types of portable and light weight penetrometers are available which include the Dynamic Cone Penetrometer (DCP) developed by (Sowers and Hedges, 1966), and the German Light weight Penetrometers (LRS5, LRS 10 – DIN 4094 Part 2). The LRS 10 Penetrometer was used in this study to characterize the soil condition on highway alignment. Both the DCP and LRS 10 are easy to deploy and dismantle. In its original form the DCP consists of 6.8 kg steel mass falling through a height of 50.8 cm that strikes the anvil to cause penetration of a 3.8 cm diameter 45° vertex angle cone that has been seated in the bottom of a hand-augered hole. The blows required to drive the embedded cone to a depth of 450 mm have been correlated by others to N values derived from the Standard Penetration Test (SPT). Experience has shown that the DCP can be used effectively in augered holes to depths of 4.0 m to 6.0 m. In performing the tests a cumulative number of blows over the depth of penetration are recorded. There have been variations of the DCP in terms of hammer weight; an 8 kg hammer is sometimes used. The DCP is often used to evaluate or characterize pavements. It is also used to characterize the sub-grade.

The LRS 10 penetrometer

The LRS 10 is equipped with an anvil and driving rod, a 10 kg rammer, rammer fall of 50 cm, 11 sounding rods, lifting device for sounding rods, and couplings all in a box casing weighing approximately 71 kg. If soil to be investigated is not in the dense to very dense consistencies penetration test can be carried out to between 10 to 12 m. The tip cone can either be 45° or 60°. The cone tip angle of the penetrometer used in this study is the 60°. The rods are 20.0 mm in diameter. The LRS 10 acquires data in the number of blows per 10 cm penetration. DIN 4094 Part 2, (1980) gives guidance on both qualitative and quantitative interpretation of the LRS 10 readings. Qualitative interpretation factors that influence the LRS 10 reistance pattern when used in predominantly sandy soil according to “DIN4094” include skin friction, overburden and depth, ground water, and influence of layer limit. For quantitave interpretation, “DIN 4094” gives an equation that estimates relative densities of diffrent soil strata in situ. The equation is of the form

$$ {\displaystyle {I}_D} = 0.21+ 0.230\ log\ {\displaystyle {n}_{10}} $$
(1)

where n 10 is the number of blows per 10 cm. Compared to the SPT, the LRS 10 probes 10 cm thickness of soil at a time versus 30.5 cm for the SPT. It can therefore detect changes in soil consistency within shorter reaches than the SPT.

Site investigation

Penetration test

The proposed alignment partly follows the existing road. The road has an unpaved shoulder. The penetrations tests were carried out 4.0 to 5.0 m after the edge of the existing pavement. Overburden materials consisting of vegetative materials with humus were first removed until a uniform soil color was encountered or exposed. The thickness of these overburden materials varies from about 0.150 m to about 0.40 m. After the removal of the overburden, the LRS 10 was set up and used to probe the soil consistencies with depth along the proposed alignment. The Penetrometer tests were located at intervals of 250 m on either side of the alignment, alternately in most cases of the tests, but otherwise at some locations. Depth of investigation was 5.0 m to 6.0 m. The test points are designated serially starting from Test No. 1 to Test No. 55. Test no. 1 was located at 21.650 km, while Test No. 55 at 36.400 km, 00 + 000 km being a mileage post in Uyo City representing the origin of the distances along the alignment. These tests were performed during the month of March, before the full onset of the rainy season.

Soil sample collection

Disturbed soil samples were collected from trial pits dug along the alignment. They were dug to a depth of 3.0 m and are located close to the points where penetration tests were carried out. Samples were collected all along the depths as there were no visible changes in soil types or lithology.

Laboratory analysis

Samples of soil collected at different trial pit locations and depths were tested for index properties. Other tests carried out on them include: Moisture- Density relationship test using West African compaction test (modified Proctor Type A-ASTM-D1557), and California Bearing Ratio test. All tests were performed in accordance with relevant ASTM standards.

Results and discussion

Soil indices, classification, and physical properties

Sieve Analysis

Sieve analysis results for some of the soil samples are presented in Table 1 which shows soil indices and basic properties of all the sampled locations. The results showed that for most of the soil sampled, the percentage passing sieve no. 200 is less than 50 %, with values ranging between 23.8 % at 22.65 km to 45.6 % at 33.25 km, placing most of the soil in the coarse grained texture. Exception to this trend are soils sampled from kilometers 26.85, 27.25, 34.75, and 35.00 along the alignment where the percentages passing no. 200 sieve are 57, 58.2, 54.8, and 56.4 % respectively. Some grain size analysis results are presented in Fig. 1.

Table 1 Soil indices, properties, and classification
Full size table
Fig. 1
figure 1

Grain size analysis of some of the soil samples along highway alignments

Full size image

Atterberg limits

Some Atterberg Limits results are also presented in Table 1. Liquid Limit values ranged from 30.0 to 55.9 %, while Plastic Limits values are between 16.9 and 36.6 %. Plasticity Index values range from a minimum 8.2 % to a maximum value of 24.3 %, indicating low to moderately swelling potential as presented in Fig. 2 which is a modified Casagrande plasticity chart. This also shows plots of soil domains.

Fig. 2
figure 2

Plots of soil domains along pavement route alignment

Full size image

Soil classification

Soil laboratory analyses results of Atterberg Limits with sieve analysis places the soil investigated in quite a wide range of soil types. Based on the American Association of State Highways and Transportation Officials (AASHTO) soil classification systems soil along the proposed alignment includes A–2–4, A–2–6, A–2–7, A–7–5, A–7–6, and A–6 (at one location). Unified Soil Classification System (USCS) places the dominant soils along the alignment into Clayey Sand (SC), Silty Sand (SM), and combinations of the two (SC-SM). There is occurrence of Inorganic Silt (MH and ML) and OH in four locations along the alignment, at chainages 26.85 km, 27.25 km, 34.75 km, and 35 km. These results are presented in Table 1.

Natural moisture content and liquidity index

The natural moisture content of the soil samples is from 11.2 to 16.9 % while liquidity index is from −1.5 to −0.3. Typical values are between −1.0 to +1.0 (Peck et al. 1974).

Soil strength

Compaction tests

The compaction test results for some of the samples tested are presented in Table 1. The results of the tests which were carried out using the modified AASHTO method showed that the value for Maximum Dry Density (MDD) for soil ranges from 1630.0 kg/m3 at distance 27.25 km to 1990 kg/m3 at 22.65 km; and the optimum moisture content (OMC) ranges from 9.4 to 20.5 % at 22.45 km and 27.25 km respectively.

Laboratory California bearing ratio test

California Bearing Ratio (CBR) values range from 15.5 % at location 33.05 km to 29.5 % at 22.65 km . Test points with liquidity index values of −1.2 and below have exceptionally low values of CBR. These samples were collected at distances 23.65, 24.05, 24.45, 26.25, 26.65, 26.85, 27.25, 30.05 and 38.40 kilometers along the alignment route. Exceptions to this trend are soils at locations 22.05 km and 25.45 km with a CBR value of 29 %. All the locations listed with Liquidity Index values of −1.2 and below have CBR values between 15 and 20.5 % representing values that are the lowest in the CBR range of values. These soils are mostly SM, SC-SM, ML, OH, and MH types. Soil in these locations are said to be sensitive, that is, they lose strength when disturbed. Das (1983). Some of these results are presented in Table 1. Physically the soil samples collected are fine to medium grain, dark yellowish in color and semi-solid in consistency.

Ground water

Static ground water table was not encountered in any of the trial pits dug for the collection of soil samples throughout the length of the alignment.

Penetration resistance, penetration index, and relative density

The LRS 10 Penetrometer used for the penetration tests have similar dimensions of the critical parts, and the operation is similar to the Dynamic Cone Penetrometer (DCP). The main difference is the size of the hammer which is slightly heavier in LRS 10, being 10 kg. While the standard DCP delivers 45.472 Joules of energy with a hammer of 8 kg falling through 575 mm, the LRS 10 delivers 49 Joules falling through 500 mm, resulting in a difference in energy of 3.528 Joules. This translates to the estimation of Penetration Index values (the key parameter in penetrometer analysis) in error of about 7.75 % (3.528/45.472) on the conservative side.

The LRS 10 readings for the first 1.50 m are therefore interpreted like DCP readings. The interpretation was carried out using the UK DCP software version 3.1 (Done and Samuel 2006). The software analyzes the LRS 10 readings and identifies layers of different strengths as measured by Penetration Index and lists the CBR values for each layer and the depth of each layer from the ground surface. The program output is a chart that shows a plot of these identified layers. A Typical chart is displayed in Fig. 3. The software plots cumulative numbers of blows versus depth, and identifies layers based on points falling on the same straight line where the slope is different from the next straight line joining another set of points. The slope of such line of locus is the Penetration Index. The software uses an equation that relates Penetration Index with CBR. The equation was proposed by Transport and Road Research Laboratory (1990), and is given as

Fig. 3
figure 3

Typical output chart from UK DCP software version 3.1 software

Full size image
$$ Lo{g}_{10\kern0.75em }(CBR)=2.48-1.057Lo{g}_{10}(penrate) $$
(2)

Where

pen rate = penetration index.

The software places a limit on the depth for which it gives reliable interpretation of data at 1.50 m. At depths beyond this, the interpretation is susceptible to errors due to higher skin friction and increase in overburden stress. However, DIN 4049, Part 2 (1980) is recourse for further interpretation of data from depths beyond the 1.50 m.

The LRS 10 readings from the 0.00 m to 6.0 m depths were converted to relative bedding density (in situ) using equation (1) above. This equation results in numerical values which can be used to classify the relative densities of the in situ density into very loose, loose, medium dense, dense, and very dense. This classification is presented in Tables 2, 3 and 4 for some of the tests.

Table 2 Sample of LRS10 penetration tests results, DCP analysis results showing soil layering based on penetration index values (Zone 1)
Full size table
Table 3 Sample of LRS10 penetration tests results, DCP analysis results showing soil layering based on penetration index values (Zone 2)
Full size table
Table 4 Sample of LRS10 penetration tests results, DCP analysis results showing soil layering based on penetration index values (Zone 3)
Full size table

Three trends are discernible from the results of this classification of in situ density of the soil along the proposed alignment.

  1. a)

    Zone 1-Surficial layer of loose soil.

From the results of the LRS 10 and computations of the in situ relative densities there are some test point locations in which there is loose soil within the first 0.30 m to 0.45 m from the ground surface. It will be necessary to remove these loose deposits when construction begins. The test point locations having this surficial loose cover are listed in Table 5, and typical field tests data for such zones are presented in Table 2.

Table 5 Classification of soil zones along alignment route based on In- situ relative density
Full size table
  1. b)

    Zone 2 - Loose soil at both surface and intermediate depths

In this zone the loose layer is both within 0.30 m - 0.45 m of the surface and occurs in thick sequence of at least 0.30 m within 2.30 m of the surface with only two exceptions in which the loose band occurs up to 3.30 m depth. These are at test points 25 (28.450 km) and 47 (35.250 km). These bands of loose soil occur to within 2.90 m depth at test points 43 (34.250 km) and 51 (37.120 km).

Test point locations characterized with this trend of loose band of soil are listed in Table 5. This zone is the most dominant.

  1. c)

    Zone 3 - Loose soil completely absent

These locations do not have loose soil present at all, both at the surface and at depth. Test locations with such characteristics are listed in Table 5, and Table 4 presents samples of such test locations.

The loose soil in zones 1 and 2 within 0.0 m to 0.45 m of the existing ground surface can be removed as part of ground stripping and removal of vegetative layer. However, beyond 0.45 m depth there is need to increase the density of the soil up to 2.90 m depth. This can be achieved by compaction of the soil with impact or polygonal compactor machines. These machines have compactive effort effect that can penetrate depth up to 2.10 m or more (Kloubert 2009; Jaksa et al. 2012), and will consolidate the soil up to the required depth and thereby strengthen the soil.

Correlations

Correlation of CBR with relative density index

Using the CBR values obtained by the software UKDCP from the analysis of penetration values (mm/blow) in the 0.0 m to 1.50 m depth range, a correlation between field CBR values and relative density gives the following relationship:

$$ CBR=0.1548\kern.3em ex{p}^{11.793\kern.2em {I}_D} $$
(3)

Where I D . = Field relative density as calculated using equation (1) above that was proposed by German DIN 4049 Part 2, (1980). Figure 4 is a plot showing the relationship.

Fig. 4
figure 4

Field California Bearing Ratio values plot with in situ relative density

Full size image

The values of relative densities used in obtaining equation 3 above are from loose (0.255) to medium range (0.476)). There was no value in the dense or very dense region. According to DIN 4049, Part 2 (1980), the maximum In situ relative density that can be sounded by LRS 10 is 0.5 in soil in which the ground water table is not encountered during exploration, and approximately 0.55 where ground water is encountered.

Correlation of penetration index with relative density values

The layering results from the software analysis are based on soil thickness with the same Penetration Index values. The layering indicated by relative density value analysis is based on consistencies of soil layers, which are: very loose, loose, medium, dense, very dense consistencies. Numerical ranges of values are associated with each of the states of consistency, and these are found in a standard table. The layering by the two approaches is consistent in some cases (Test 7 presented in Table 4 are examples) and are widely different in some cases (Test 12, Test 19 presented in Table 2 are examples). The latter situation indicates that the range of numerical values represented by loose and medium consistencies are too wide and could mask soil having different strengths. From Table 4, Test 24, while relative density indicates ‘medium’consistency from 0.0 m to 6.0 m, penetration index classification shows penetration index values from 5.26 to 13.79 within the depth range 0.0 m to 1.50 m, identifying five layers. The penetration index classification has better resolution than the relative density consistencies. A plot between penetration index and relative density is presented in Fig. 5 and the correlation coefficient between the two is indicated. The correlation equation is:

Fig. 5
figure 5

Plot of penetration index versus in situ relative density values

Full size image
$$ DCPI=1135.7\ \mathrm{e}\mathrm{x}\mathrm{p}\left(-,10.83,, {I}_D\right) $$
(4)

Where DCPI = Dynamic Cone Penetration Index, mm/blow

I D = relative density values.

Subgrade stiffness-resilient and Young’s modulus estimation

The Resilient Modulus (MR) is a measure of sub-grade material stiffness. A material’s resilient modulus is an estimate of its modulus of elasticity (E). While the modulus of elasticity is stress divided by strain for a slowly applied load, resilient modulus is stress divided by strain for rapidly applied loads (Pavement interactive, 2007). During a resilient modulus test, a very “small” permanent strain accompanies each load cycle; this is absent in Elastic modulus test (Irwin 2009). The laboratory determination of resilient modulus is quite involved, hence other methods of estimating its value from other soil properties that can be readily determined was explored.

A number of relationships are available in literature on the estimation of resilient modulus using the Penetration Index of the DCP and CBR laboratory data. For DCP, Pen (1990); De Beer and van der Merwe (1991); Lockwood et al. (1992); Lee et al. (1997a); George and Uddin (2000); and Chen et al. (2005), all gave relationships that allow estimation of resilient modulus from Penetration Index values. Some specified the type of soil (coarse or fine grained); the expression is applicable to. The values of resilient modulus for the test point locations obtained with expressions proposed by Lockwood et al. (1992); Lee et al. (1997b); and George and Uddin (2000), are presented in Tables 6 and 7. The equations utilized are:

Table 6 Resilient modulus values of the natural subgrade computed using penetration index values
Full size table
Table 7 Resilient modulus values computed with penetration index values of 30 mm/blow and above
Full size table
$$ {\mathrm{M}}_{\mathrm{R}}=1{0}^{3.04758-1.06166\ \log\ \left[\mathrm{DCPI}\right]} $$
(5)

By Lockwood et al.(1992), for any type of soil.

$$ {\mathrm{M}}_{\mathrm{R}}=-3279\mathrm{DCPI}+114100 $$
(6)

By Lee et al. (1997a), for fine grained soil and,

$$ {\mathrm{M}}_{\mathrm{R}}=235.3\times {\mathrm{DCPI}}^{-0.475} $$
(7)

by George and Uddin (2000), for coarse grained soil.

in these relationships,

MR = Resilient Modulus in kPa in equation (1), and MPa in equations (6) and (7)

DCPI = Dynamic Cone Penetration Test Index.

The Penetration Index values used were from the results of DCP software analysis which analyses the DCP data from zero to 1.5 m depth. A further refinement of the result was made by using the highest value of Penetration Index (lowest strength) obtained in the range of 0.30 m to 1.50 m depth for each test location. The top 0.30 m were excluded since this may likely be removed during earthworks stripping operations.

Values of resilient modulus obtained by the three expressions for DCPI values of 30 mm/blow and above, (30.7 mm/blow, 33.33 mm/blow, 50 mm/blow and 100 mm/blow) correlates poorly. There are wide differences in values of the MR computed by these expressions. Test positions that have this trend are Test Nos. 2, 6, 10, 25, 28, 29, 30, 33, 34, 38, 43, 46, & 53. The values at these test locations are presented in Table 7. For DCPI values below 30 mm/blow the difference in resilient modulus values between the lowest and the highest values obtained by all of these methods, in most cases, is not more than 10 MPa. The results are presented in Table 6. Exceptions to this are values with DCPI of 25 mm/blow and 26.67 mm/blow. The values for these exceptions are not as widely dispersed as the values for those with DCPI of 30 mm/blow and above. The above indicated that for soil in ‘loose’ state the evaluation of resilient modulus in situ with light weight dynamic penetrometer is not reliable. However, with respect to DCPI values that are less than 30 mm/blow, the fact that the values estimated from the three expressions are comparable (maximum difference in values of 10 MPa), though the soil type to which one of the expressions is applicable, is fine grained indicates that:

  • the values obtained are near the true value of in situ modulus of the soil at their depths of occurrence, and

  • the expressions represent reliable methods of determining true sub-grade stiffness as represented by resilient modulus values for the types of soils encountered in this study.

Plots of resilient modulus values computed with the three expressions and the field CBR is presented in Fig. 6. While Lee et al. (1997b) expression correlates poorly with a linear correlation coefficient of 0.541, George and Uddin (2000), and Lockwood et al. (1992) have linear correlation coefficient (R) of 0.970 and 0.988 respectively showing good correlation. Lee et al. (1997a), however, correlates better when fitted with logarithmic model with ‘R’ equals to 0.883. The fact that data points by all the three are not widely scattered from one another and are close towards the origin indicates better estimates of the true values for resilient modulus by the three expressions. Typical values of in situ bulk unit weight of soil on the alignment are 18.41 kN/m3 and 18.21 kN/m3; this gives overburden pressure at 1.50 m depth of about 27.615 kPa. This could be taken to represent the confining stress for the resilient modulus computed in Tables 6 and 7.

Fig. 6
figure 6

Plot of estimated resilient modulus from dynamic cone penetration index values versus field California Bearing Ratio values

Full size image

Laboratory determination of resilient modulus values as noted above is not readily available in the study area but can be estimated from the CBR test which is familiar and widely used. Young’s modulus values were estimated for each test sample using the Laboratory CBR data and the stress at the CBR value. In the calculation the penetration of the CBR plunger machine was considered as deformation at a given force value. Plunger area for the machine used was 19.4 cm2. Young’s modulus values (E) were computed for 2.5 mm. 5.0 mm, 7.5 mm and 10.0 mm deformations and associated strain level. The maximum of the Young’s modulus values are taken, mostly at 5 mm penetration. The results for the laboratory CBR values are presented in Table 8. Back calculations were made to estimate penetration index values (DCPI) for each laboratory CBR value. These were carried out using equation 2 above; the DCPI values obtained are that of the tested soil at OMC. Since laboratory resilient modulus are usually determined at or near optimum water content for a given soil sample, estimation of the soil resilient modulus were then computed using the three relationships above represented by equations 5, 6, and 7. These are presented also in Table 8.

Table 8 Resilient modulus computed from back calculated penetration index from laboratory CBR values
Full size table

The values of resilient modulus computed are now regressed with the values of Young’s modulus obtained from the laboratory CBR data. Figure 7, presents the correlation results. The correlations equations by which resilient modulus can be obtained from Young’s modulus derived from laboratory data are:

Fig. 7
figure 7

Regression of resilient modulus values with Young’s modulus values computed from laboratory CBR

Full size image
$$ {M}_R=33.048 In\left({E}_R\right)-62.776 $$
(8)
$$ {M}_R=1.065{E}_R+12.69 $$
(9)
$$ {M}_R=0.440{E}_R + 43.384 $$
(10)

Where M R = Resilient modulus in (MPa), and E R = Young’s modulus in (MPa), estimated from laboratory CBR at optimum moisture content with value ranges between 40.53 and 83.68 MPa).

Correlation coefficients for each of the equations are 0.850, 0.855, and 0.853 respectively.

These expressions are proposed for coarse grained soils.

Anochie-Boateng et al. (2010) reports laboratory values of resilient modulus between 56.5 MPa and 99.2 MPa. These soils were tested at optimum moisture content and soaked for four days. The soils were from the sub-grades of an airport runway, taxiways and apron. The soils classify as MH, ML, SM, CL, and SC. The soaked (96 h) CBR values vary from as low as 2.5 % up to 18.7 %. Mohammad et al. (2007) reports the range of resilient modulus values for some soil types, namely A–4, A–6, A–7–5, and A–7–6. These values are 35 to 69 MPa, 28 to 97 MPa, 7 to 96 MPa, and 21 to 62 MPa respectively. Resilient modulus values obtained in this study from the back-calculated DCPI from the laboratory CBR (24 h soaked) values ranged from 51 MPa to a maximum of 100 MPa. With similarity in soil types and index properties of the Mohammad et al. (2007) study; the ranges of resilient modulus values obtained in this study compare well with values from some previous works.

Laboratory versus field values

For safe design it is the practice to use the lowest value of relevant design parameters. The lowest value of both the field CBR and resilient modulus were obtained. Compared to laboratory CBR, the field CBR values within 1.50 m of the ground surface, in some cases, are higher, and otherwise in others. The maximum difference was about 19.5 % at test location 28, excluding the locations with field CBR values of 55 and 83 %. The range of the field CBR is larger (2 to 30 %) than the range of laboratory one which is from 15.5 to 29.5 %. The resilient modulus values follow a similar trend with values ranging between zero to 108 MPa (except 130, 202 and 304 MPa) as range for field, and 50 to 100 MPa as laboratory values. Since laboratory CBR and resilient modulus are to serve as control for field earthworks operations this implies that the desired field CBR values can easily be achieved and, in some cases, surpass desired values. With respect to resilient modulus, the field values were the least available in situ; for those locations with values less than 50 MPa (the minimum laboratory value) the field resilient modulus values have to be brought up to this value at those locations. The above strongly suggests that field resilient modulus values obtained using the LRS 10 penetrometer adequately estimate sub-grade stiffness required for design purposes for a given pavement.

Conclusions

The highway alignment is characterized by sandy and silty soil types within 0 to 6.0 m depth investigated. These soil types are Clayey Sand, SC (A-2-6), Silty Sand SM(A-2-7), and a combination of the two, SC- SM (A-6). There is occurrence of MH, OH and ML (A-7-5) soil types within stretches of the alignment as shown by four test points. These soil types are consistent with the geology of the area. Based on soil consistency as indicated by relative density, a light weight Penetrometer was able to characterize the alignment in to three zones in both depth and areal extent, adequate for the proposed pavement. AASHTO classification rate as “good”, a significant stretch of the sub-grade alignment based on both the in situ CBR and the lowest laboratory CBR values.

The sub-grade stiffness as indicated by both in situ tests with LRS 10 and estimated resilient modulus values for a reasonable portion of the highway alignment is adequate; however, improvement of the relatively ‘loose’ by deep compaction sections will need to be required.

Correlation was established for the type of soil encountered in this study between CBR and relative density, and between penetration index and relative density. Similarly, a relationship was established between resilient modulus and Young’s modulus. The latter can easily be evaluated from laboratory CBR tests.