YEAR 2006

AN APPROACH TO DEVELOP GRANULAR SUB BASE MIXES USING LOCALLY AVAILABLE MATERIALS
By A. U. Ravi Shankar1, Suresha. S. N1 & Anil Kumar. G2

CONTENTS
Pages
1. Introduction ......... ......... ......... 1
2. Literature Review ......... ......... ......... 2
3. Experimental Investigations ......... ......... ......... 3
4. Results and Discussion ......... ......... ......... 4
5. Conclusions ......... ......... ......... 8

ABSTRACT
The stringent specifications and non-availability of the materials satisfying the specifications of the Granular Sub-Base (GSB) course is one of the problems in road construction industry. Although, the provision is there for using locally available materials in GSB layer, most of these fail to satisfy the Atterberg limits/gradation/soaked CBR requirements. This paper summarises the investigation carried out with an objective of developing granular sub-base mixes using the locally available materials and to arrive at a mix-design approach. The study was carried out using the locally available materials namely, lateritic soil, crusher-run-dust, and tile bats, along with river sand and crushed granite aggregates. The laboratory tests like Grain Size Distribution, Atterberg Limits, Modified Proctor Density, and soaked CBR tests were carried out for the basic materials, blended soil, and for GSB mixes. Based on the investigations, a design approach is established for developing GSB mixes using locally available materials.
Key Words: GSB; lateritic soil; crusher-run-dust; tile bats; air voids; soaked CBR.

1. INTRODUCTION
Provisions of Granular Sub-Base (GSB) layer became necessary in the current construction practice of all the national highway projects in India when pavements are constructed on weak subgrade. Sub-base is an intermediate layer between subgrade and granular base course. The function of this layer is to act as a drainage layer for the pavement to avoid excessive wetting and weakening of subgrade. In strength it is more superior as compared to the subgrade. Various materials and techniques are used for construction of sub-base course. National Rural Roads Development Agency (NRRDA) greatly encourages the utilization of locally available and industrial waste materials in the construction works of rural road pavements, as it also solves the problem of disposal of huge amount of industrial waste1.
The specification2 suggested by the Ministry of Road Transport and Highways (MORT&H), has the provision for using two-types of GSB mixes, i.e. close-graded and coarse graded type. In each type, three different gradations are recommended based on the maximum nominal size of the aggregates. In case of rural roads, Indian Roads Congress (IRC)3 recommends coarse graded GSB layer with minimum soaked CBR of 15 per cent. In addition to the gradation and strength requirements, GSB mix has to satisfy the Atterberg limits i.e. the Liquid Limit (LL) and Plasticity Index (PI) should be less than 25 and 6 per cent respectively for the material passing through 425-micron sieve2,3.

1.1. Objectives
In developing GSB mixes, the materials, like, natural sand, moorum, gravel, crushed stone, or combination of these are being used. But, in certain regions, like, Konkan belt of India, these materials are not available abundantly to meet the requirements of road construction activity. To satisfy the local needs, the present investigation has been carried out with the following objectives:
· To alter the properties of the locally available materials by proper blending;
· To proportion the blended soil with coarser materials to meet the requirements of GSB gradations;
· To arrive at a design approach for developing GSB mixes.

1.2. Scope
In the present investigation locally available materials namely; Lateritic Soil (LS); Crusher-Run-Dust (CRD); and broken Tile Bats (TB) were used along with River Sand (RS) and Crushed Granite Aggregates (CGA). The physical and strength properties of the individual materials are evaluated as per the standard test methods. Blending of the above materials in different combination was carried out to in order to achieve the GSB specifications. Proportioning of materials is carried out as per Rothfuch’s graphical method4 and by computational trail and error method (using MS Excel Worksheet). The properties like, Atterberg limits; grain size distribution; moisture-density relationship under heavy compaction, and soaked CBR values are determined for the proportioned GSB mixes.

2. LITERATURE REVIEW
Pavement structures generally consist of four layers: subgrade; sub-base course; unbounded/bounded base course, and wearing surface. The purpose of providing the sub-base course is to drain out the water, which can be extremely deleterious to the life of the pavement. The main sources of free water in pavement systems5 include water infiltrating through cracks in the pavement; water entering longitudinal pavement/shoulder joints; seepage water from ditches and medians; and high ground water table. In flexible pavements, water with fine material can also be pumped out causing enlargement of void spaces in the pavement base6. Excessive fines in the gradation make aggregate particles to float in the matrix resulting in low permeability with low stability7. Maximum CBR is achieved at fines content between 6 per cent and 14 per cent for granular sub-base materials, and CBR decreases from dense to open gradations8.

While using locally available materials, like, lateritic soils, care should be taken to avoid the problems due to more fines content, higher Atterberg limits, and moisture susceptible character. Laterites are generally deep brown in colour. They are porous and spongy and hence absorb water and become soft. Care must be taken when using this material in humid and cold climate. These are found capping the hills of central and western India and also along west coast and east coast9. In the west coast, Laterites are found extensively along the Karnataka and Kerala coast. In Karnataka, Laterites occur along the coastal tract of South Canara and North Canara districts. Because of their hardening property on exposure and high compressive strength, they are quite extensively used as building materials. The term ‘Laterite’ first appeared in scientific literature nearly two hundred years ago. The word ‘Laterite’ was first introduced by Buchanan10 to denote a building material used in the mountain regions of Malabar, India. Various studies using different approaches have been carried out on the hardening characteristics and properties of lateritic soils and reported in myriad technical papers11,12,13.

The use of lateritic gravel in airfield pavement sub-bases reported satisfactory performance in Brazil14. But in semi-rigid region of Peru, airfields built with lateritic gravels failed to meet the conventional airfield base and sub-base criteria15. Rollings et al.16 briefly discussed over the development and engineering issues associated with the compaction of fine-grained soils in humid regions, difficulties associated with the laboratory assessment of certain tropical soils, road construction on expansive black-cotton soils, contaminant problems in arid regions, and the use of non-standard materials including duricrusts, lateritic gravels, and coral.

Lateritic soils constitute an important group of residual soils of India, covering an area of about 1,00,000 sq. km9. They are found mainly on the western and eastern coasts over large areas and in small quantities in the southern and eastern states of India. Water plays a major role when it enters the pavement structure, causing a progressive decrease in the density and strength and an increase in the plasticity and possible swell or shrinkage13. Lateritic soils can be stabilised by various means and can be utilized as base and sub-base course in pavement17.

The studies carried out at Central Road Research Institute, India; reveal that, locally available low-grade materials like dhandla soil-gravel, moorum, kankar and stabilized soil can be used effectively and economically in pavement layers of rural roads18. Berthelot et al.19 carried out investigation on cement modification of granular base and sub-base materials using Tri-axial Frequency Sweep Characterization and concluded with recommendation of adding cement to marginal quality aggregates for better performance. Carlos20 reported that the brickbats in base courses showed good record of use and performance in Bangladesh. Brick macadam bases become denser under traffic and develop high strength, while being flexible and insensitive to moisture at the same time.

Kazuhiko and Kuboi21 based on their studies recommended that the effective utilization of the waste rock powder resources as a construction material could solve the problem of environmental pollution (air and water) and disposal in Japan.

Research & Development studies and successful field demonstration projects have proved that waste materials like fly ash, iron and steel industry slags, municipal waste, rice husk ash, marble slurry dust, recycled concrete etc. can be used for construction of roads3.

3. EXPERIMENTAL INVESTIGATIONS

3.1. Materials

The lateritic soil was collected from the road construction site near to Surathkal. The physical and strength properties of the lateritic soil are tabulated in Table 1. It is clear from the results that the lateritic soil contains more than 10 per cent of fines passing 75-micron and liquid limit and plasticity index are more than 25 and 6 per cent, respectively, which are undesirable for the sub base course. The soaked CBR value is quite different under heavy and light compaction tests. It is clear from the above facts, that the lateritic soil needs some modification to its gradation and Atterberg limits before using it as granular sub base.

Non-plastic materials; namely crusher-run-dust and river sand were used to blend with lateritic soil to modify mainly its Atterberg limits. Crusher-run-dust was obtained from local stone quarry and sand from Nandini riverbed. The properties of crusher-run-dust and river sand are presented in Table 1.

To meet the requirements of coarser size fractions in GSB gradation, materials like tile bats and crushed granite aggregates were used. The waste tile bats were procured from the local tile industry. To get the required gradation, these bats were crushed in to small pieces by hand hammer. Crushed granite aggregates were obtained from the near by granite stone quarry. The physical requirements, namely, 10 per cent fines value and water absorption values were determined for the coarser size fractions. The 10 per cent fine values (as per BS: 812-Part-111) of crushed granite aggregate and tile bats were more than 50 kN. The water absorption value (as per IS: 2386-Part-3) of crushed granite aggregates and tile bats was 0.15 per cent and 19.1 per cent, respectively. The loss in the soundness test (as per IS: 383) carried out on tile bats were within the acceptable limits. The other properties of the tile bats and crushed granite aggregates are presented in Table 2. The major chemical composition of the tile bats, tested at Tiles Research & Development Centre, National Institute of Technology Karnataka, Surathkal is tabulated in the Table 3. Depending upon the gradation of GSB, different sizes of tile bats and crushed granite aggregates are used, viz. 75/50/25/10 mm down size. In addition to this, fractions properties of the crusher-run-dust significantly affected the Atterberg limits of lateritic soil.


By addition of 40 per cent of crusher-run-dust in total mix, the liquid limit and plasticity index values come down to 23.9 per cent and 3.87 per cent respectively, which is within the acceptable limits of GSB specifications (LL <25 per cent and PI < 6 per cent).

Fig. 2. Variation of Atterberg limits with per cent of CRD/RS
Effect of Sand: When lateritic soil was blended with 50 per cent of sand liquid limit reduced from 38per cent to 30.3per cent, and plasticity index from 12 per cent to 7.0per cent. The reduction in index properties is relatively low compared to crusher-run-dust; which may be because of silt fractions in the sand, and due to relatively low specific gravity compared to crusher-run-dust resulting in more surface area. The liquid limit and plasticity index of blend reduced to 24.1 per cent and 4.2 per cent when lateritic soil was blended with 70 per cent of sand.

Fig. 3. Moisture-density relationship of blended lateritic soil
4.1.2. Moisture-Density Relation of Blended Lateritic Soil: The moisture-density relationship of the lateritic soil blended with different percentage of crusher-run-dust/river sand was studied under IS Heavy compaction method. From Fig.3, it is observed that the moisture-density curves for blended soils are towards the dry side of the moisture-density curve of lateritic soil, and it may be due to coarser grains in blended soils. The moisture-density curves of crusher-run-dust blended lateritic soils are relatively more peaked when compared to the lateritic soils blended with sand.
When lateritic soil is blended with crusher-run-dust of 30 to 50 per cent by weight of total mix, MDD varies between 2149 and 2170 kg/m3 and OMC gradually decreases from 8.7

per cent to 7.9 per cent. In case of lateritic soil blended with sand, MDD decreases with addition of sand. Addition of 60per cent of sand to lateritic soil resulted in a decrease in the MDD from 2180 kg/m3 to 2154 kg/m3 and a reduction in OMC from 10.4 per cent to 8.3 per cent. Further addition of sand i.e. at 80 per cent, MDD decreases to 2008 kg/m3, but OMC increases from 8.3 per cent to 10.2 per cent, this variation trend is shown in Fig.4.

4.2.1. Proportioning of Material: Proportioning of materials was carried out to meet the close-graded and open graded gradation requirements of GSB. At this stage sand is not used in preparing GSB mixes, as it requires more than 70 per cent replacement to get the specified Atterberg limits. The blending operation is carried out using lateritic soil, crusher-run-dust and coarser aggregates (tile bats/ crushed granite aggregate). The proportioning of all these material was done based on the Rothfuch’s graphical method/ trial and error method. To meet the specification requirements for the material passing 425-micron, the minimum quantity of crusher-run-dust is kept 30 per cent and above. To meet the requirements of coarser size fractions in GSB gradations, tile bats/ crushed granite aggregates of different nominal sizes were used in the blend of lateritic and crusher-run-dust. The proportions and gradation of GSB mixes for different gradations are shown in Tables 4 and 5.

The specified gradations for close-graded GSB mixes contain higher finer fractions when compared to open-graded GSB mixes. This has reflected in proportioning of the mixes, where in the close-graded GSB mixes up to 30 per cent of lateritic soil was used. But, in open-graded GSB mixes, except for grading III, the use of lateritic soil is limited to 23 per cent. The quantity of crusher-run-dust is kept 20 per cent of total mix for close-graded GSB mixes, 12 per cent for open-graded GSB mixes except Grading III. In the mix of lateritic soil and crusher-run-dust, the proportion of latter varies in between 33 and 40 per cent depending upon the gradation of GSB mix. The coarser fractions were substituted either by tile bats or crushed granite aggregates; the gradation and proportion of coarser fractions are as shown in Tables 4 and 5.

   

In the combined gradations for different proportions of lateritic soil, crusher-run-dust, and tile bats/crushed granite aggregates were within the specified gradation limits of different GSB mixes. In the combined gradation of the close-graded GSB mix of grading II, the materials corresponding to the sieve size of 4.75 mm are having higher coarser fractions and it is very close to the lower limit of the specified gradation. But, in open-graded GSB mix of grading I, the materials passing 4.75 mm contains relatively higher finer fractions and it is very close to the upper limit of specified gradation limit. To avoid this problem, there was a need to suitably modify the gradation of crusher-run-dust.
4.2.2. Moisture-Density Relationship: The relationship between the dry density and moisture content was obtained based on the IS heavy compaction test. The materials passing 20 mm sieve were used for the test, and the coarser size fractions (i.e. more than 20 mm) are suitably adjusted by lower size fractions. The moisture-density relations of GSB mixes with tile bats and with crushed granite aggregates are shown in Fig.5 and 6, respectively.

Fig. 5. Moisture-density relationship for LS + QD + TB blended GSB

Fig. 6. Moisture-density relationship for LS + QD + CGA blended GSB

The MDD and OMC for the different gradations of open-graded GSB mixes obtained by blending lateritic soil with crusher-run-dust and tile bats are in the range of 1831 to 1889 kg/m3 and 14 to 15 per cent respectively. The MDD and OMC for close graded GSB mixes are in the range of 1915 to 2065 kg/m3 and 8.3 to 11.6 per cent. The lower maximum dry density and higher optimum moisture content recorded for the open graded mixes are because of the presence of tile bats, i.e. lower specific gravity and more water absorption character. In open-graded mixes, the amounts of coarser fractions are relatively more when compared to close-graded mixes. In the GSB mixes blended with lateritic soil, crusher-run-dust and crushed granite aggregates, the MDD and OMC recorded for different gradations of open and close-graded types are in the range of 2148 to 2210 kg/m3 and 7.0 per cent to 10.0 per cent respectively. The variations in MDD and OMC values in open and closed graded GSB mixes with crushed granite aggregates is not that significant when compared to GSB mixes with tile bats.

4.2.3. Strength Property of GSB mixes: The strength of different GSB mixes was evaluated in terms of soaked CBR values. For grading II and III materials, the soaked CBR values were evaluated at the density and moisture content likely to develop in equilibrium condition, which shall be taken as density relating to a uniform air voids content of 5 per cent2. The density corresponding to the air voids of 5 per cent (approximately) and moisture content towards dry side of the optimum is arrived by using equation (1). The GSB mixes were compacted as per IS heavy compaction method and the soaked CBR values observed were more than 30 per cent. However, it is observed that the soaked CBR value of original lateritic soil corresponding to MDD and OMC under heavy compaction itself is more than 30 per cent. The properties like MDD, OMC, air voids and corresponding density and moisture content are tabulated in Table 6.

 

4.3. Discrete GSB Gradations for Locally Available Materials
The investigations carried out in the above two stages correspond to materials of single gradation. In reality, the gradations of individual materials used in the investigation will not be exact in the actual practice. Hence, the investigation carried out needs to be generalized with respect to gradation of materials. In this regard, an attempt has been made to analyze the possible gradations for each material. Two probable gradations were considered for each material on either side of the original gradation i.e. coarser and finer gradations. Proportioning of the materials as per the changed gradations has been analyzed. The gradation of the blend is checked with specified GSB gradation. Further, the viability of the gradations with reference to specified Atterberg limits (i.e. LL <25 per cent and PI < 6 per cent) was cross-checked with Figures 1 and 2, i.e. considering the materials passing 425-micron and the corresponding index properties. All these investigations were carried out analytically and based on the previous two stage investigations. The finalized gradation limits for individual materials are shown in Table 7. These simplified gradations will help in developing GSB with locally available materials.
Table 7. Recommended Gradations for Local Materials for GSB

5. CONCLUSIONS
The investigation was aimed to develop an approach for developing GSB mixes using locally available materials namely lateritic soil, river sand, crusher-run-dust, tile bats and crushed stone aggregates. The studies were carried out in two stages and based on the results obtained from the investigations the following conclusions were made:

4.3. Discrete GSB Gradations for Locally Available Materials
The investigations carried out in the above two stages correspond to materials of single gradation. In reality, the gradations of individual materials used in the investigation will not be exact in the actual practice. Hence, the investigation carried out needs to be generalized with respect to gradation of materials. In this regard, an attempt has been made to analyze the possible gradations for each material. Two probable gradations were considered for each material on either side of the original gradation i.e. coarser and finer gradations. Proportioning of the materials as per the changed gradations has been analyzed. The gradation of the blend is checked with specified GSB gradation. Further, the viability of the gradations with reference to specified Atterberg limits (i.e. LL <25 per cent and PI < 6 per cent) was cross-checked with Figures 1 and 2, i.e. considering the materials passing 425-micron and the corresponding index properties. All these investigations were carried out analytically and based on the previous two stage investigations. The finalized gradation limits for individual materials are shown in Table 7. These simplified gradations will help in developing GSB with locally available materials.
Table 7. Recommended Gradations for Local Materials for GSB

5. CONCLUSIONS
The investigation was aimed to develop an approach for developing GSB mixes using locally available materials namely lateritic soil, river sand, crusher-run-dust, tile bats and crushed stone aggregates. The studies were carried out in two stages and based on the results obtained from the investigations the following conclusions were made:
· Modification to Original soil (lateritic soil)

o Particle size between 425 and 75-microns are relatively more in crusher-run-dust when compared to the river sand. Therefore, by adding a low percentage of crusher-run-dust to the lateritic soil the Atterberg limits of the soil were altered to suit the requirements.

o Use of river sand was restricted to the first stage of investigation, as it requires replacement of more than 70 per cent of lateritic soil to satisfy the specifications of Atterberg limits.

· Use of Tile Bats and Crushed granite aggregates in GSB Mixes

o Both tile bats and crushed granite aggregates satisfied the requirement of 10 per cent fines value.

o Though the soaked tile bats were used in the GSB mixes, the OMC for these mixes were recorded higher when compared to the OMC of the lateritic soil.

o The soaked CBR values of different GSB gradations corresponding to the density and moisture content at 5per cent air voids were more than 30per cent.·

· Approach to develop GSB mixes

o Investigation should be carried out in two stages i.e. modification of original soil (when its LL > 25 and PI>6) to satisfy the Atterberg limits, and blending the modified soil with coarser fractions to meet the desired GSB gradation.

o Modification of lateritic soil should be carried out with non-plastic materials (modifiers).

o The fractions passing through 425-micron should be higher in modifiers when compared to original soil. This will significantly contribute in reducing the Atterberg’s limits with lower quantity of modifier.

o Coarser size fractions can be substituted by using tile bats/crushed granite aggregates but care must be taken to determine the OMC in case of materials having high water absorption property like tile bats.

o The probable variation in grain size distribution of individual materials should be considered and for these variations, combined gradation should be checked to meet the GSB gradations.

ACKNOWLEDGEMENTS

The support and encouragement given by the National Rural Roads Development Agency (NRRDA) to carry out the investigation is thankfully acknowledged.

REFERENCES

1. Prem S. Tripathy, and S.N. Mukherjee, “Perspectives on Bulk Use of Fly ash,” CFRI Golden Jubilee Monograph, Allied Publishers, New Delhi, 1997, pp.117-118.

2. Specifications for Road and Bridge Works, Ministry of Road Transport and Highways, Indian Roads Congress, New Delhi, 2001, pp. 101-105.

3. Rural Roads Manual, Special Publication Number: 20, Indian Roads Congress, New Delhi, 2002.

4. Khanna, S.K., and C.E.G Justo, “ Highway Engineering.” 7th Ed., Nem Chand & Bros., Roorkee, 1991, pp. 681-682.

5. Baumgardner, H. R., “Overview of Permeable Bases. Materials: Performance and Prevention of Deficiencies and Failures, 1992.” Materials Engineering Congress, Atlanta, GA, 1992, pp. 275-287.

6. Randolph W. B., Heydinger G. A., Gupta D. J., Jiangeng Cai, Edward Stienhauser, Quinglu Xie, “Permeability and Stability of Base and Sub-base Materials.” Final Report No. FHWA/OH 2000/017, submitted to the Ohio Department of Transportation, Department of Civil Engineering, University of Toledo, Ohio, 2000.

7. Thornton, S. I., and Elliott, P. R., “Fines Content of Granular Base Material.” Final Report TRC-8703, submitted to Arkansas Highway and Transportation Department, Department of Civil Engineering, University of Arkansas, Fayetteville, Arkansas, 1988.

8. David J. White, Pavana Vennapusa, and Charles T. Jahren, “Determination of Optimum Base Characteristics for Pavements.” Final Report, Iowa DOT Project TR-482, Department of Civil, Construction and Environmental Engineering, Ames, IA, May 2004.

9. Gopal Ranjan and A.S.R. Rao, “Basic and Applied Soil Mechanics.” Willey Eastern Limited, New Delhi, 1991, pp. 8-9.

10. Buchanan, F., A Journey from Madras through the countries of Mysore, Canara, and Malabar, Vol. 3, East India Company, London, United Kingdom, 1807.