Geotechnical Investigation and Effect of Moisture Content on Subgrade CBR Values; Arbaminch-Chencha Existing Road; Ethiopia

In the flexible pavements sub-grade is the undermost layer for pavement structure constructed on it and used to resist traffic load. Hence, subgrade geotechnical behaviors are considered as measuring parameters of the pavement design and performance. But, here Arbaminch-Chencha road is constructed as gravel road twelve years so far without any geotechnical investigation. The aim of this research is geotechnical investigation and effect of moisture content on CBR of the subgrade soil of ArbaminchChencha road. This is to classify and characterize this subgrade according to its geotechnical properties, to identify effect of moisture on subgrade CBR for selected three soil classes. To achieve these objectives, seventeen representative test samples collected along the existing road. Laboratory investigations performed for in-situ moisture content, specific gravity, grain size analysis, Atterberg’s limit test for all pit samples. And then soils are classified. After classification of soils; laboratory moisture – density relationship and CBR values are determined. The effect of moisture on its CBR was observed for specified soil classes. For all tests the apparatus and the procedures used for analysis were done according to ASTM and AASHTO standards. According to AASHTO soil classification, Arbaminch-Chencha road subgrade soils classified as coarse and fine grained soil; but dominated by fine grained soil class. As per observed, laboratory maximum dry density of this soil ranges from 1.36 gm./cc to 2.11 gm./cc and OMC range from 10 % to 30%. And four day soaked CBR Values are ranges from 3 % to 49 % under specification of 95% MDD. Effect of moisture revealed as, CBR value reduced 4% to 28% at dry side whereas 25% to 59% at wet side of OMC and MDD of specified soil classes. Saturation after 48 to 96 hours, CBR reduced from 17% to 30% whereas 96 to 144hours reduced 7% to 20%. But, continuous and linear reduction observed in finer soil classes.


INTRODUCTION
In the highway pavement structure, sub-grade is a naturally consolidated and/or constructed layer to resist traffic load which provides a suitable foundation for the pavement layer constructed on it. Hence, sufficient information on soil class, characteristics and mechanical properties (basically California Bearing Ratio) of foundation material is very important for engineering constructions; such as road projects. Arbaminch-Chencha road is found in Gamo Gofa Zone, in southern Ethiopia; which is constructed twelve years before (in 2005). This road is currently gravel paved with having heavy traffic flow. Hence, it gives large transportation purpose for different areas around and far from the road. Areas using this road are to and from Arbaminch-Dorze, Chencha, Ezo, Ditta, Dara Malo (Waca), Abassa, and to other currently developing small towns and villages. This encourages the social and economic development activities of these areas by transporting an industrial raw products (textiles) and wide agricultural products (fruits; Apple, Apple mango etc.), potato, maize, animals for meat) of the Arbaminch-Chencha highlands. Although, with having heavy traffic flow and transportation purpose by small vehicles to heavy trucks, it is till now a low class gravel road constructed so far without any geotechnical investigation on its subgrade soil. As information collected from Gamo-Gofa road and transport office, Arbaminch-Chencha road is constructed under this office as a client and they informed for this research work as no soil investigation conducted during that time at all. But now a day this road is under investigation by ERA for upgrading it to asphalt concrete level because of its heavy transportation purpose and the rapid development of community there. Since, every activity requires quality road as much as possible to transport different products as discussed in above paragraph. Hence, the present research is directed on geotechnical investigation and effect of moisture content on CBR values on subgrade soil of Arbaminch-Chencha existing road.

STATEMENT OF THE PROBLEM
Investigation of soil and soil materials is so important since all infrastructure construction projects (like roads) use naturally occurring soils and soil Martials as the basic foundation as well as construction materials. Unlike manmade materials, the properties of these soil are highly variable and a function of the complex natural processes that occurred in the geologic past. And are heterogeneous, nonlinear material, and typically anisotropic instead of being isotropic. As a consequence, constructions like pavement structures are facing problems related with soils and soil materials available in the project site, whose properties are often unknown and of variable quality. Arbaminch-Chencha road was constructed without any geotechnical investigation and has been giving transportation purpose for the community with continuous maintenance twice per year on most part of the road. These maintenances are conducting just before and after rainy season. This is because of the capacity of this road is not equivalent to its purpose; since it was gravel paved. Most of this road section is found in different climate zones according to Ethiopian climate zoning. Particularly, there is high precipitation and low temperature is common at the end section of the study area; Dorze -Chencha section. Rain water stays for more than a week in poorly constructed side ditch during rainy seasons in this section. Not only climate problem, but also there is shallow ground water table near to the surface about 2 to 3m. This ground water "(BONO)" is used for day to day purpose of the area. Due to this, subgrade strength (CBR value) can be affected by long term moisture saturation. But when the road comes to the starting point of this research, climate is different from the above section; in which precipitation is low with high temperature. The subgrade strength owing to its inconsistency or variable nature poses a challenge to come up with a perfect design of pavement on it. Since, the subgrade is always subjected to change in its moisture content due to precipitation, capillary action, flood or abrupt rise of water table. Change in moisture content causes change in the subgrade strength. And it becomes quite essential for an engineer to understand the exact nature of dependence of subgrade strength on moisture content. Therefore, this research is helpful to know the geotechnical behavior and effect of moisture content on CBR values of the subgrade soil was studied.

General Objective
The general objective of this research is geotechnical investigation for subgrade and effect of moisture content on its subgrade CBR values of Arbaminch-Chencha existing road.

Specific Objectives
The specific objectives of this research work are: To classify and characterize subgrade soils found along Arbaminch-Chencha road alignment; for road construction purpose. By using in-situ moisture content, specific gravity, grain size analysis test and atterberg's limt tests To determine laboratory compaction characteristics and subgrade CBR values of subgrade in the Arbaminch-Chencha road and To experiment and evaluate the effect of moisture content on subgrade CBR in this road alignment 4. METHODOLOGY To achieve the objectives, seventeen (17) sampling pits areas selected and their geographical location is determined in field using GPS reading. From these sample pits, disturbed subgrade soil samples were excavated to a maximum depth of 1.5 meters and collected for laboratory investigation (as per ERA site investigation manual 2013). Laboratory tests were conducted for natural moisture content, specific gravity, Atterberg's limit test, grain size analysis on all seventeen samples. And moisture-density relationship, CBR and CBR swelling under OMC is performed on ten specified pit samples. Lastly, the effect of moisture content on CBR values is evaluated for different soil classes by (1) by varying moisture content during preparation of CBR specimens before (dry side) and after (wet side) OMC; (2) by extending saturation period beyond 96hours (4days) to 144hours (6days). During any activity for this research, special attentions paid for the sack of tropical residual soil properties; even though there is no given standard to investigate such type of soils.
For 17 pit samples of subgrade soil along road stretch, pit chainage was started from Chano (pit -1; stationed 9+160 by considering 0+000 station at Arbaminch town) and proceeded to Chencha town (pit -17; stationed 27+200) as shown in figure 4.1. Chano town exists 9km from Arbaminch to east on the Addis Ababa Arbaminch main highway. Location of pits taken some distance offset right hand side(RHS) or left hand side(LHS) from centerline of road, since road is under use. So that is impossible to dig pits just at centerline of road which was justified in the following table 4.1. Selection of pits for various tests was made on the following criteria: (1) variability of subgrade soil along the roadway, (2) consideration on samples taken from these variable locations are representative for overall subgrade along this road. And also (3) distance of minimum of 2 to 3 km distance between these pits as much as possible as per ERA pavement design manual; vloume1, 2013. Manual recommends taking strength test samples with limit of 2 to 5km to have economical design. Since, it is too costly to design subgrade CBR with in shorter distance less than 2km.

Effect of Moisture Content on CBR Values of Subgrade
The variation of moisture in soils has great adverse impact on quality and performance of structures constructed on it. Since, increase in moisture content in substructure material decreases the engineering quality of soil; like load bearing capacity. The variation of moisture content in soil and soil materials may be developed most likely from climate change as discussed in chapter two. Thus, Arbamich-chencha road has two different climate zones; these are (1) road near Arbaminch (about 9 km from 19km total road) which has more hot time and more evaporation than precipitation per year and (2) the second part is road near Chencha (about 10 km from 19km total road) part in which there is more precipitation and low temperature, shallow ground water table (2m to 3m). Hence, moisture fluctuation and its effect on subgrade bearing capacity were expected, particularly during rainy seasons. This research work was conducted to study moisture effect on subgrade CBR value in two ways: (1) Preparing CBR specimens with water contents at dry side and wet side of OMC and (2) preparing CBR specimens at OMC but varying the degree of saturation (period of saturation).
Three soil classes (pit-1; A-2-4 soil class, pit-5; A-7-5 soil class and pit-14; A-4 soil class) were selected for the above tests. These selections were made on the bases of soil classification, CBR values and to cover over all study area along the road.

Effect of Moisture Content on CBR Values by drying and wetting
Materials used in this section were the same as used in normal (conventional CBR test) CBR test in section 4.1. Air dried soil materials passing 19mm prepared and modified if material has larger size than 19mm.
The variable parameter selected in this case was moisture content. Two moisture contents were selected on dry side of OMC and two on wet side as discussed in chapter five and other parameters were kept constant as much as possible.
So that, four CBR specimens were prepared; two CBR specimens at dry side and two on wet side. Each CBR specimen from these soil classes was compacted by energy of 56 blows per layer to make variables other than moisture content constant. Dry density and moisture contents were determined. Then after, these remolded CBR specimens were soaked for 96 hours with surcharge load of 4.56kg. Initial and final readings for percentage of swelling were taken before and after soaked for determination of swelling.

Effect of Moisture on CBR values by Period of Saturation
The second way performed here to evaluate effect of moisture content on CBR values of subgrade was extended saturation period. Commonly, it is popular to use 96 hour's soaked CBR values of subgrade for the design of pavement structures. This was accepted by considering 96 hours saturated is the worst and CBR value at this period is the lowest and no effect could be observed.
However, this condition may difficult to accept in some conditions such as at area in which there is high rainfall and low temperature soaking CBR specimens for 96 hours can't saturate fully whereas at areas in which there is low rainfall and high temperature soaking CBR specimens for 96 hours not logical. Since, in the first case moisture can affect CBR values by gradual saturation for longer time than 96 hours. And for the second case 96 hours soaked CBR values are not economical because of no such effect could be expected.
Thus, for this research the effect of moisture content was studied by varying saturation period 0, 2(48 hours), 4(96 hours) and 6(144 hours) days on CBR specimens from those selected soil samples.
To do so, four CBR specimens from above three soil classes were prepared by OMC and compacted in 56 blows per layer. Materials used here were the same material as materials used in section 4.2.1. These specimens (unsoaked, 2 days soaked, 4 days soaked and 6 days soaked) were soaked for corresponding saturation periods under surcharge load of 4.56kg. Dry density, moisture content and percentage of swelling all were determined the same way as procedures taken in conventional CBR test AASTO T193.
But moisture content was taken from three points (top about 2 cm below from the top edge of the mold, middle and bottom about 2cm above from the bottom edge of the mold) from single CBR specimen and determined as per ASTM D2216/AASHTO T265. Then the average of three is used for density determination of CBR specimen for each soil.
CBR values and percentage of swelling for every unsoaked, soaked for 2, 4 and 6 days specimens were determined as described. And overall observed, analyzed, tables and figures are collected in section 5.2.

Geotechnical Investigation
In this chapter, investigation results discussed are: in-situ moisture content, specific gravity, soil classification tests such as sieve analysis and Atterberg's limit (Liquid limit, Plastic limit, and Plasticity Index value) are presented. Soils are classified. Then laboratory moisture-density characteristics and CBR test under normal conditions (at OMC and MDD) were explained. The effect of moisture content observed under moisture content other than OMC and extended degree of saturation period results are presented here. All the laboratory results are summarized in tables and figure in the following sections.

In-Situ Moisture Content
Methods of determination of in-situ moisture contents in samples collected from 17 pits are presented in section 4.1. The results obtained from these tests are presented in table 5.1. As their results indicated that there was variation of moisture content observed from pit to pit (table 5.1) along the road. The variation was most likely due to the location of pits with respect to their elevation and weather condition. This was not only the reason but also soil by itself varies from pit to pit along the road. In-situ moisture content was relatively higher at higher elevations near Chencha and the reverse was also true. Pit samples containing fine soil content had the greater in-situ moisture than those containing lesser fine soil or more coarser particles.
Soils here under study for this research, the in-situ moisture content obtained from laboratory undisturbed and disturbed samples by ASTM D2216/ASHTO T-265 range in between 10 % to 44% except pit 16 has 73%( which was relatively higher and close to its liquid limit). The average of all test results was 30.15 % which is more likely in range of silty and clayey soil. Since, most part of soils in the study area was dominated by fine grained soil class.

Specific Gravity
Method to determine the specific gravity is described in  The results of specific gravity vary from soil class to soil class. Hence, its values of the study area varies from 2.62-2.75 which is in the range of typical specific gravity values of gravel, sand, silt and clay soils. The lower values represent coarser soils and the higher values are for finer soils.

Grain Size Analysis
The grain size analysis test performed to have group of soil grains with in their corresponding sizes which are helpful for plasticity characteristics and soil classification. Here table 5.2 is developed as summary for seventeen pit samples from grain size observations and analyzed results.
The grain size analysis results indicate that the dominant proportion of particle in the study area is fine soil (silt and clay) of average percentage of 44.17% and the second dominancy by sand size of average of 39.07% and the last was gravel of average of 16.46%. As observed in table above, soils are finer and finer when it approaching to higher elevation along the road. Therefore, finer soils are found around Chencha town.

Atterberg Limits test results
As described in chapter four, results of Atterberg limits (liquid limit, plastic limit) are determined according to AASHTO T-89 and T-90 standard test methods. Typical liquid limit test flow curve is given in Figure 5.1 for sample pit-1.  Soils are grouped under both granular materials (35% or less passed on 0.075mm sieve) and silty-clay materials (> 35% passed by 0.075mm) general class. These general classified soil classes are divided in to sub-classification subgroups as A-1 to A-7 and corresponding soil descriptions are also given in Table 5.4.  From overall, it is revealed that soils in Arbaminch-Chencha existing road is dominated by fine grained soil classes. This is 59% fine grained (10 from 17 pits) whereas 41% is coarse grained soils (7 from 17 pits).

Moisture-density relationship
Natural subgrade compaction is common and purposeful to have compaction characteristics of it. It is required to densify the existing subgrade materials for overlaying pavement structures. Results of moisture density relationship were also used to in the preparation of CBR test specimens. From this test, the required parameters are maximum dry density (MDD g/cc) and optimum moisture content (OMC %) which were determined from different densities achieved by different moisture contents. With having different densities and moisture contents density-moisture curve was plotted as curve plotted for pit-16 as shown in figure 5.3 and MDD and OMC were obtained as follow.   Now, to obtain CBR values at specification of 93%, 95% and 97% MDD; figure 5.5 are plotted (a) for CBR to dry density and (b) for percentage of swelling to dry density. These curves are very important to have desired specification by moving on curves as per AASHTO and ERA standards. CBR and swell values were determined by using equation found in corresponding curves; y-values as CBR or swell values whereas x-values for dry density. By the similar way overall CBR and swelling results for 10 pit samples were presented in table 5.9 with including corrected load-penetration curves.  Clay soil For these three soil classes; CBR values were determined for four CBR specimens from each soil classes. These CBR specimens; two from dry side and other two from wet side as described in chapter four.   The effect of moisture on subgrade CBR values in this section was presented in figure 5.8. The variation of (a) CBR with saturation period and (b) CBR with moisture was observed from chart drawn here in figure 5.8. And combined chart was present in figure 5.9. By the same manner the effect was observed for all three soil classes in appendix-I and summarized here in table 5.13. As effect observed; reduction of CBR varies with saturation period for all different soil classes. But the reduction is different in different soil class. Thus, for finer soils (A-7-5(30)) saturation water increases linearly whereas for coarser soils (A-2-4(0)) saturation water increases highly at initial stage up to saturation time 2 days and gradually decreases as emphasized in figure 5.9. Table 5.14 presents the variation of CBR due to saturation period as discussed above and shows CBR values respect to 4 days soaked CBR (which is taken from table 5.9 for 56 blows).  Here, results showed that the reduction of CBR varies with saturation period for different soil classes. For example: A-2-4 soil class has 82%, 70%, 65% for 2, 4 and 6 days saturated CBR values with respect to unsoaked CBR value. Whereas A-7-5 soil class has 77%, 62% and 50% with respect to unsoaked CBR value and 82%, 63% and 53% with respect to unsoaked CBR for soil class of A-4.
From this, coarser soil classes (A-2-4 and A-4) reduced more just before 48hours saturation and decreases slightly after 96hours whereas finer soil classes (A-7-5) reduced CBR gradually up to 144hours and continuous somewhat more degree than above soil classes. But, the effect is more in fine grained soils than coarse grained soils. Hence, unsoaked CBR values of soil classes vary from 43% (coarser soil class) to 61% (for finer soils) higher than 4 days soaked values. Here in figure 5.10; the effect of moisture content by saturation period was nicely described for 2 to 4 and 4 to 6 days. 1) The in-situ moisture content of this subgrade ranges from 10 to 44% whereas the specific gravity is ranges from 2.62 to 2.75. 2) Subgrade soil classified by AASHTO M145 classification system as: A-1-a, A-2-4, A-2-5, A-2-6, A-3, A-4, A-5, A-7-5, and which was dominated by fine grained soil class. 3) From Moisture-density relationship, optimum moisture content(OMC%) ranges from 10% to 30% whereas maximum dry density (MDD g/cc) ranges from 1.36 to 2.11 g/cc. for general concept dry density decreases when the road comes to Chencha from Arbaminch. 4) Some of the pits in the road are good for both subgrade and subbase material for road construction purpose. For example, pit 3 and 8 having PI of 3% and NP, 95% MDD CBR value of 37% and 47 % respectively. Hence, these subgrade materials are good for construction of sub base. 5) Effect of moisture content on CBR was different for different soil classes before and after OMC. As observed, CBR value reduced 4% to 28% at dry side whereas 25% to 59 % at wet side of OMC and MDD of specified soil classes. This indicates that, wet side reduced CBR values 21% to 31% more than dry side. 6) Saturation after 48 to 96 hours, CBR reduced from 17% to 30% whereas 96 to 144hours reduced 7% to 20%. But, more reduction observed in finer soil classes. Therefore, saturation beyond conventional soaking period is significant in the areas in which there is high precipitation; low evaporation and shallow ground water is expected