1. Introduction

Reinforcement of marginal or substandard soils becomes necessary in engineering applications when the available strength falls short of requirement and yet the soils must be used. One way to improve soil properties for pavement construction is to reinforce the material with geosynthetics. One such material which has gained increasing acceptance in soil reinforcement is geogrid. A geogrid is a geosynthetic material consisting of connected parallel sets of tensile ribs with apertures of sufficient size to allow strike-through of surrounding soil, stone, or other geotechnical materials. There is substantial agreement on the benefits that can be achieved from the inclusion of geogrid within a pavement system, however, the extent of the improvement is in relative disagreement ^[1].

When geogrids are used as reinforcement for pavement layers, both the soil and geogrid contribute to the load distribution mechanism ^[1]. It is known that geogrid provides aggregate interlock which stabilizes the soil and prevents lateral movement of soil aggregate particles ^[1]. However, the degree to which the grading and plasticity characteristics of the soil influence strength development is not well reported.

This study investigated the interaction between the plasticity index (PI) and gradation properties of lateritic gravel within a soil-geogrid reinforced composite structure for a pavement. The aim was to contribute to the understanding of how material properties affect the California Bearing Ratio (CBR) enhancement for different placement locations of geogrid reinforcement within a pavement structure. This study, therefore, attempted to answer the following questions: To what extent does the plasticity index affect strength development in a geogrid-reinforced soil? How does the amount of coarse aggregates in a geogrid-reinforced soil influence strength enhancement? Does the depth of geogrid placement in a layer influence strength enhancement? Answers to these questions are required to improve our understanding of the design of aggregate-geogrid composites in road pavement. More importantly, the use of geogrid reinforcement to enhance the CBR of pavement layer materials could prove very beneficial in reducing the volume of borrow materials rejected for being marginal in quality. Also the inclusion of geogrid in pavements will help provide stiffer pavements which could resist plastic deformation better. This paper present initial results of the ongoing experimental study.

2. Literature Review

Placement of unbound pavement soil/aggregate layer materials in road construction is generally expensive, therefore, it is often important to find ways of minimizing the aggregate layer thickness for a given service life of road. Where pavement layer materials have substandard strength, one way to improve on the material’s response to load is to incorporate geosynthetic in the structure as strength reinforcement. Alternatively, incorporating a geosynthetic in a road pavement structure can increase the service life for a given aggregate layer thickness. In view of this, parameters necessary for the design and incorporation of geogrid in pavement structures are very necessary. So far, research on geosynthetic reinforcement has tended to concentrate mostly on advancing the geosynthetic product rather than the behavior of the soil when reinforced. This makes research which provides understanding of the behavior of natural gravel reinforced with geogrid very imperative. It seems that in the past three decades, more effort has been devoted to perfecting the manufacturing processes of geogrids rather than optimizing their structure and properties to improve their interaction with soil ^[1].

Lateritic soils abound in countries with tropical weather and are favoured over crushed rock as pavement layer material because of their relatively low cost. According to Gidigasu ^[2], the engineering behavior of laterites depend mainly on the genesis and degree of weathering. The geotechnical properties of laterites most relevant to their engineering performance are particle size distribution, Atterberg limits, strength of the coarse particles, compaction characteristics and bearing strength ^[3]. Particle size distribution is important with respect to compactability and finish ^[4]. Lateritic soils with high plasticity give indication of high clay content and are likely to pose problems under wet conditions. Strength of compacted pavement layer materials is usually evaluated using California Bearing Ratio (CBR), although some standards may require the crushing strength of the coarse aggregates as well.

The use of combinations of plasticity and grading indices is also prevalent. The specification that has probably been the most successfully applied to the greatest length of roads is that used in Brazil which also makes use of the CBR for lateritic gravels. The Standard Brazilian specifications for base courses for roads (DNIT 141/2010) has requirements for particle size distribution (typical Fuller-type requirements), Atterberg limits (PI< 6%), dust ratio, CBR (60% for less than 5 million axles and 80% for more) and Los Angeles Abrasion loss (max of 55%) ^[1].

According to Koerner ^[5], the first use of fabrics in reinforcing roads was attempted by the South Carolina Highway Department in 1926. A heavy cotton fabric was placed on a primed earth base, hot asphalt was applied to the fabric, and a thin layer of sand was placed on the asphalt. The Department published the results of this work in 1935 for a number of field trials. Until the fabric deteriorated, cracking, raveling and localized road failures reduced. This project was certainly the forerunner of the separation and reinforcement functions of geosynthetic materials, as we know them today ^[5].

Geogrids consist of heavy strands of plastic materials arranged as longitudinal and transverse elements to outline a uniformly distributed and relatively large and grid-like array of apertures in the resulting sheet. These apertures allow direct contact between soil particles on either side of the sheet ^[5]. Geogrids are characterized by integrally connected elements with in-plane apertures (openings) uniformly distributed between the elements. The apertures allow the soil to fill the space between the elements, thereby increasing soil interaction with the geogrid and ensuring unrestricted vertical drainage. All these applications are not only in highway, but also in railroad track construction and rehabilitation ^[6]. Geogrids have been used successfully in pavement layer studies. Montanelli et al. ^[7] reported of CBR increase when geogrid was placed between a sand subgrade and a gravel base course. Gosavi et al. ^[8] reported of CBR increase for a soil reinforced with mixed geogrid woven fabric. Work by Naeini et al. ^[9] indicated that using a geogrid as layer reinforcement in soil samples with different plasticity caused considerable increase in the CBR value compared with the unreinforced soils in both soaked and unsoaked conditions.

In the absence of a reinforcing aggregate layer, a subgrade soil subjected to repeated traffic loading, will experience rutting due to accumulated plastic deformation and/or bearing capacity failure. However, if there is an overlying soil/aggregate layer, any rutting observed at the surface of the aggregate layer will essentially be a combination of decrease in aggregate layer thickness at the location of the ruts, as a result of traffic compaction, and deformation of the subgrade soil. Decrease in thickness of the aggregate layer will also decrease the ability of the aggregate layer to distribute the load on the subgrade soil, thereby, increasing the deformation of the subgrade soil further ^[1].

Two mechanisms account for aggregate layer deterioration; interpenetration (mixing) of aggregate and subgrade soil, and displacement of aggregate within the aggregate layer ^[1]. Interpenetration of aggregate layer and subgrade soil results from repeated loading. It manifests in two ways: downward movement of aggregate (i.e. loss of aggregate into the subgrade), and upward movement of fine particles from the subgrade soil (i.e. intrusion of fine subgrade soil particles into the aggregate layer). The loss of aggregate into the subgrade decreases the thickness of the aggregate layer, which also decreases its ability to distribute the traffic loads. The intrusion of fine subgrade particles into the aggregate alters the mechanical properties of the aggregate layer and makes the layer more likely to deform and less able to distribute traffic loads.

In addition to load distribution, geogrid reinforcement can contribute to load support through a mechanism called the “tensioned membrane effect”. It is to be noted that the tensioned membrane effect had, in the past, been wrongly taken as the main mechanism governing the performance of unpaved roads ^[1]. The tensioned membrane effect contributes to decrease in the stress induced in the subgrade soil under wheel loads by transferring part of the load to lateral zones (i.e. away from the wheels).

When a pavement is loaded, the geogrid deforms in sympathy with the deflection bowl created, assumes a concave shape and enters into a state of tension. The resultants (which are oriented upward) of the geogrid tensions on each side of the concave shape contribute to wheel support. These resultants are balanced by downward resultants associated with the convex shape of the geogrid away from the wheels ^[1]. From the foregoing, it is clear that the tensioned membrane effect requires deep rutting to be effective and works only with channelized traffic (i.e., if traffic keeps deepening the same ruts) ^[1].

3. Materials and Method

This study used lateritic soil collected from Suame, a suburb of Kumasi, Ghana. Different percentages (10% and 20%) by weight of clay powder were used as admixture for changing the gradation and plasticity index of the original soil sample. The modified soils were identified as S10 and S20 to denote addition of 10% and 20% clay, respectively, while the original soil was identified as S0 to indicate no addition of clay powder.

The experimental work involved index property tests (specific gravity, gradation, Atterberg limits, and compaction characteristics) and CBR tests, on the original and modified soils. The compaction tests on the original and modified soils were compacted in accordance with the procedures given in ASTM-D 1557 ^[10], using modified effort to establish their optimum water content and dry densities. For CBR tests, samples of the original and modified soils were compacted at their corresponding optimum water contents, with and without geogrid reinforcement. A biaxial geogrid type GX 30x30 was used as the reinforcing material. Properties of the geogrid were obtained from the manufacturer’s specifications. The aperture size was 30mm.

Reinforcement involved placement of geogrid material as the samples were being compacted in layers in accordance with the following schedule:

Single layer reinforcement

• Top of Layer 1

• Top of Layer 2

Double layer reinforcement

• Top of Layers 2 and 4

The strengths of the reinforced and unreinforced samples in terms of CBR were determined for soaked and unsoaked conditions in accordance with ASTM-D 1883 ^[11].

4. Results and Discussion

Table 1 contains data on the index properties of the study soils in terms of compaction characteristics, (optimum moisture content-OMC, dry density-ρ_d), liquid limit (LL), plastic limit (PL), plasticity index (PI), specific gravity (G_s), the fine fractions (less than 2mm) and coarse fractions (greater than 2mm). Based on the fractional component data in the table, the coarse to fine fraction ratio () for soils S0, S10 and S20 were computed to be 3.3, 2.8 and 1.6, respectively. In terms of soil plasticity, Sample S20 was the most plastic while S0 was the least. This was expected as the original sample was progressively modified by adding increasing amounts of clayey fines to alter the gradation and also plasticity. The data on the index properties do not seem to show much differences between the soils in terms of compaction characteristics (i.e., ρ_d and OMC) and liquid limit but show some appreciable differences in terms the plasticity indices.

Table 1. Index properties of study soils

Download as

PowerPoint Slide

Larger image(png format)

• Tables index

Veiw figure

View current table in a new window

View next table

Download as

• PPT
  PowerPoint Slide
• PNG
  Larger image(png format)

Veiw figureFigures index

•  NEW
  View larger figure in new window

View next figure

Figure 1. Particle size distribution curves of study soils and MRH specification envelope

Distinct differences in the soils, however, existed in terms of gradation as shown by Figure 1, which is a plot of the particle size distributions superimposed on the gradation envelope for pavement layer materials specified by the Ministry of Roads and Highways (MRH), Ghana ^[12].

It is seen that whereas Sample S0 lies within the MRH specification envelope, parts of S10 and S20 are outside. Thus, S0 satisfies MRH gradation specification for pavement material whereas S10 would be classified as marginal material; S20 may be rejected for lying completely outside the MRH envelope. On the basis of plasticity index, both S10 and S20 will not satisfy MRH requirements for pavement layers (MRH requires PI<18). In spite of these limitations, the study went ahead to investigate the materials because good quality lateritic soils for road construction are becoming increasingly scarce in the country and an understanding of how those of marginal quality could be improved for use is considered helpful.

The results of the unsoaked/soaked CBR tests on the unreinforced and reinforced soils are presented in Tables 2 and 3, respectively. By introducing in this paper the parameter strength ratio (SR) as the ratio of the CBR of the reinforced sample to the CBR of the unreinforced sample, Table 4 may be generated from the test data for soaked conditions. Even though data in this study were limited, there is a clear indication that the soils (with and without reinforcement) generally lost strength of the order of 50% in terms of CBR when soaked. For the particular case of sample S20 (with and without reinforcement), which inexplicably had very high CBR values in the unsoaked state, the loss in strength under soaked conditions was dramatic (well over 80%).

Table 2. Unsoaked CBR values in percent

Download as

PowerPoint Slide

Larger image(png format)

• Tables index

Veiw figure

View current table in a new window

View previous table

View next table

Table 3. Soaked CBR values in percent

Download as

PowerPoint Slide

Larger image(png format)

• Tables index

Veiw figure

View current table in a new window

View previous table

View next table

Table 4. Strength Ratio (SR) for soaked conditions

Download as

PowerPoint Slide

Larger image(png format)

• Tables index

Veiw figure

View current table in a new window

View previous table

Based on the SR values in Table 4, the gain in strength (measured by CBR) of the soils across the various geogrid placement types in the soaked condition varied between 9% and 73% over the strength of the unreinforced, with the less plastic soils tending to gain more in reinforcement. As is evident in the tables, two-layer reinforcements generally did not appear to have significant edge over single-layer reinforcements whether under unsoaked or soaked conditions. Even though it is not succinctly clear from the limited test data, it does appear that Layer 3 reinforcement tended to result in only marginal enhancement in strength over that of Layer 2 reinforcement in two of the three cases.

The apparent effect of plasticity and particle size distribution on geogrid reinforcement was explored through a correlation between SR and the Coarse to Fine Fraction Ratio (CFR). Figure 2 is the outcome of such correlation for the different geogrid placements.

Download as

• PPT
  PowerPoint Slide
• PNG
  Larger image(png format)

Veiw figureFigures index

•  NEW
  View larger figure in new window
• PREV
  View previous figure

View next figure

Figure 2. Correlation between SR and CFR

From the graph, the following correlations were established in respect of geogrid placement on top of Layer 2, Layer 3, and Layer 2+4, respectively;

(1)

(2)

(3)

where,

SR=Strength ratio

CFR=Coarse to fine fraction ratio.

It is clear from the above relationships that for all reinforcement placements, SR increased as the coarse to fine fraction ratio (CFR) increased. The high R²values are indicative of a very strong correlation although it is recognized that the data set is very much limited. Increasing CFR value is generally synonymous with decreasing plasticity. In the case of this study, the loads applied were probably not large enough to cause significant strain in the geogrid, therefore, the tensile membrane effect of the geogrid was probably not called into full play. Rather, it is suspected that the geogrid was probably more effective in aiding aggregate interlocking within the aperture. Since the geogrid type, aperture size and maximum particle size were the same for all the test samples, it could be said that the factors governing strength development were more likely due to a match between the geogrid and the aggregate in the geogrid-aggregate composite layer created. Giroud ^[1] has posited that if there is perfect match between the geogrid and the aggregate, the composite layer would have optimum properties.

As indicated in the preceding section, geogrid-aggregate interaction should be mobilized with very small displacement. Therefore, the geogrid ultimate tensile strength plays no role until the reinforced structure undergoes deep rutting associated with large deformation of the aggregate layer and the subgrade soil ^{[13, 14, 15]}; the geosynthetic then provides the tensioned membrane effect. Based on the relationships developed in this paper, it is seen that the higher the coarse aggregate percentage in a soil (low plasticity index), the higher the CBR increment.

We posit that because the geogrid aggregate interlocking is enhanced as the ratio of the coarse to fine aggregate percentage increases, soil particle displacement then increases with high fines content and, therefore, high PI for all locations of geogrid placement. Thus, clearly, plasticity appears to be detrimental to geogrid reinforcement. This is because soils with high plasticity tend to have high clay content which is synonymous with low soil friction and low particle interlock.

Download as

• PPT
  PowerPoint Slide
• PNG
  Larger image(png format)

Veiw figureFigures index

•  NEW
  View larger figure in new window
• PREV
  View previous figure

Figure 3. Correlation between SR and PCA/PI

A plot of the Strength ratio versus the ratio of percentage of coarse particles (PCA) to plasticity Index was subsequently investigated. Figure 3 indicates least of the R² to be 0.75 confirming the inverse relationship of the strength ratio has with PI and the direct linear correlation between strength ratio and percentage of coarse aggregates in a geogrid-soil composite material. With the inclusion of geogrid in a soil sample, the increase in the CBR strength varies linearly as the percentage of coarse particles for the same plasticity index. Also, the inclusion of two layers of geogrid (placement at top of layer 2 and layer 4 simultaneously) increases particle confinement more than single layer occurrence for all locations investigated. From the foregoing results we posit that the rate of increase of strength is directly proportional to the percentage of coarse aggregates in the soil and varies inversely as the plasticity index for the soil being stabilized with geogrid. The degree of enhancement depend on the location of the geogrid.

5. Conclusion

Geogrid reinforcement is practiced to improve the strength characteristics of substandard or marginal soils whether as foundation soils or pavement layer materials. In the past, research on geogrid reinforcement has tended to focus largely on improving the properties of the geogrid with little attention to the influence of soil properties on geogrid performance. This study investigated the effects of plasticity index and gradation properties of lateritic soils as well as geogrid placement location on strength enhancement within a soil-geogrid-reinforced composite for a pavement. Reinforcement involved incorporating one and two layers of geogrid into the lateritic soil at different depths during the laboratory compaction. The effect of reinforcement, plasticity index, and grading on strength enhancement of the compacted material was measured by CBR for soaked and unsoaked conditions. The results indicated that as soil plasticity increased due to increased fines fraction, the CBR decreased. This means that the proportion of coarse aggregate was important for strength enhancement andthat particle interlocking becomes less effective as soil plasticity increases in geogrid-reinforced compoistes. In addition, the CBR of the reinforced soil increased for all geogrid placement locations. Two-layer geogrid reinforcement only marginally improved strength over single layer reinforcement placed close to mid-depth (from the top) of the compacted sample. This seems to suggest that single-layer geogrid reinforcement may suffice in some situations provided the placement depth is appropriate.

References

[1]	Giroud, J.P. (2009). “An assessment of the use of geogrids in unpaved roads and unpaved areas”, Proceedings of the Jubilee Symposium on Polymer Geogrid Reinforcement, Institution of Civil Engineers, London, UK, pp. 23-36.
	In article
[2]	Gidigasu, M. D. (1974). Degree of weathering in the identification of laterite materials for engineering purpose-a review. Engineering Geology, 8(3):213-266.
	In article		View Article
[3]	Madu, R. M. (1977). An investigation into the geotechnical and engineering properties of some laterites of Eastern Nigeria. Engineering Geology, 11(2):101-125.
	In article		View Article
[4]	Koerner, R.M. (2005). Designing with geosynthetics: Volume 1.Retrieved from: Amazon.com [Accessed 2 June, 2014].
	In article
[5]	Bergado, D.T. and Abuel-Naga, H. M. (2005). “Tsunami devastations and reconstruction with geosynthetics”.Retrieved from: https://www.freelibrary.com [Accessed 2 March, 2014].
	In article
[6]	Olawale, A. (2011). “Use of geosynthetics in road construction.” Dissertation, Department of Civil Engineering, Federal University of Technology. Akure, Nigeria, Retrieved from: https://www.google.com/ googlebooks.
	In article
[7]	Motanelli, F., Zhao, A., and Rimoldi, P. (1997). “Geosynthetics-reinforced pavement system: Testing and design.” Proceedings of Geosynthetics, 97, pp. 549-604.
	In article
[8]	Gosavi, G. and Patil, K. (2009). “Improvement of properties of black cotton soil subgrade through synthetic reinforcement: Effect of plasticity index and reinforcement on CBR value of soft clay”. International Journal of Civil Engineering, 7(2).
	In article
[9]	Naeini, S.A., and Moayed, R. (2009). “Effect of plasticity index and reinforcement on CBR value of soft clay”. International Journal of Civil Engineering; 7(2): 124-130.
	In article
[10]	ASTM D-1557 Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Modified Effort.
	In article
[11]	ASTM D1883ASTM Standards Tests for CBR of Laboratory Compacted soils.
	In article
[12]	Ministry of Roads and Highways, Specifications for Roadworks and Bridges 2006, Government of the Republic of Ghana.
	In article
[13]	ASTM. (1994). ASTM Standards and other specifications and test methods on the Quality Assurance of landfill linear systems.
	In article
[14]	Giroud, J.P. & Han, J. (2004). Design Method for Geogrid-Reinforced Unpaved Roads. I. Development of Design Method. ASCE Journal of Geotechnical and Geoenvironmental Engineering, 130(8).
	In article		View Article
[15]	Giroud, J.P. & Han, J. (2006). Closure to “Design Method for Geogrid-Reinforced Unpaved Roads. I: Development of Design Method.” ASCE Journal of Geotechnical and Geoenvironmental Engineering, 132(4).
	In article		View Article