DEVELOPMENT OF DESIGN PROCEDURE FOR REINFORCED FLEXIBLE PAVEMENT

Abstract

The escalating cost of materials and energy and the lack of resources available have motivated highway engineers to explore new alternatives in building roads and rehabilitating the existing ones. Reinforcement of flexible pavement is one of such alternatives. Recently, considerable interest has been generated among both highway engineers and manufacturers for using geosynthetics like geogrids and geotextiles as reinforcing materials for flexible pavement. However, absence of a well-documented design procedure for reinforced flexible pavements has resulted in low confidence in highway engineers in using these materials. Therefore, the present work was aimed at developing of a design procedure for geosynthetic-reinforced flexible pavements. Three types of soils; fine sand, sandy clay and silty clay were selected for use in this study. The granular subbase (GSB) material used in the present study was procured from a river-bed (RBM) with grading as per Ministry of Road Transport and Highways (MORTH, 2001), Government of India. Aggregates for base course were of water bound macadam (WBM) type with grading conforming to MORTH, 2001 specifications. Two types of geogrids with different stiffness were used to reinforce the pavement layers. The aperture size of both grids was chosen close to 20 mm for better interlocking, as the maximum particle size adopted for both base and subbase materials was 20 mm for consideration of sample sizes in the laboratory tests. The reinforcement was provided in the subgrade, subbase and base course layers. The asphalt course was not considered for reinforcement in this study. In the first part of the experimental program, CBR and static triaxial tests were conducted on selected materials to determine the effect of position of the geogrid and Abstract confining pressure on strength parameters of these materials. The geogrid was placed in a single layer at different positions; 20%H (specimen's height), 40%H, 60%H and 80%H from top surface of the specimen. Tests were conducted on unreinforced specimens also to compare the results. The triaxial tests were conducted at confining pressures of 0 (unconfined compression tests), 40,70,140 and 210 kPa. The results of these tests showed that the CBR value of the material increases considerably due to geogrid-reinforcement. It was the highest when the grid was placed at 100 mm depth in the CBR specimen (about 78% of the sample height from top). Also, the modulus of elasticity (E-value) of a reinforced material increases with the confining pressure and depth of the grid in the specimen. The highest value of E was obtained when grid was placed at 80% of the specimen height. The stress and strain at failure were higher for reinforced samples than for unreinforced ones. The improvement was more withlgrid of higher stiffness. The reinforcement improved the cohesion (C) of a soil. However, the effect on the angle of internal friction ($) did not bear any trend for both cohesive and cohesionless soils. In the second part of the experimental program, the plate load tests were conducted on subgrade soil using 300mm diameter plate in a test pit (2.0m x 1.0m x 0.5m) to verify the optimum position of the grid obtained from CBR and static triaxial tests. The geogrid was placed at different positions similar to the CBR and static triaxial tests and plate load tests were conducted to evaluate the modulus of subgrade reaction (k-value) for different positions of geogrid reinforcement. Simultaneously, the stress and strain distributions in the subgrade were determined using pressure cells and strain gaugesto k-used in the analytical part later. The plate load test results also indicated the highest value of modulus of subgrade reaction (K) when grid was placed at 80% of the layer depth. Further analysis of CBR, static triaxial and plate load tests indicted that for maximum benefit; the geogrid > Abstract must be placed at (72-76%) of specimen height. The height of specimen in the laboratory tests is equivalent to the depth of compacted subgrade soil or thickness of subbase/base course layer in field. The third part of the experimental program included the study of behavior of different materials under dynamic loading. Cyclic triaxial tests were conducted on reinforced materials with grid at optimum height. Cyclic loads producing the vertical stresses of 195 and 260 kPa were applied. For each vertical stress load, three confining pressures of 40,70 and 140 kPa were selected producing six different deviator stresses. The results of these tests revealed that both the permanent and resilient strains in all materials (soil, subbase and base) decrease with confining pressure but increase with the number of load cycles and deviator stress in reinforced and unreinforced conditions. Further, the resilient modulus decreases with the number of load cycles and deviator stress, and increases with the confining pressure. The results also in<fcifed the impressive enhancement in the resilient behaviour and deformation characteristics in all materials due to reinforcement. The results of the experimental program were used to develop a design procedure for geogrid-reinforced flexible pavement. A mechanistic design approach, which considers reinforcement benefits in terms of extension in service life of the pavement or reduction in base or subbase thicknesses for equivalent service life is proposed. Two important considerations, which have not been attempted in earlier research, are involved in this methodology. The first, the reinforcement is not restricted to any particular layer of a pavement. It can be used either in subgrade, subbase or base course layer. It can also be used in more than one layer depending upon the requirements. The second, the grid is considered as a part of the structural layer. The composite layer is designated by its enhanced values of strength parameters. This assumption is experimentally validated with in Abstract the obtained stresses and strains through plate load tests and also through finite element (FE) modeling using ANSYS computer code. The variation in measured and calculated stresses was 3-12% (average 7.5%) and that in strains was 2-10% (average 6%), which is quite acceptable. The FE modeling was used to predict the rutting index, the vertical strain o^t---top of the subgrade, as it was considered as the failure criterion in the present methodology. The benefits of geogrid-reinforcement were quantified in terms of traffic benefit ratio (TBR), which expresses extension of the pavement service life or in terms of reduction in the base or subbase thickness for the same service life of both reinforced and unreinforced. The lat ris termed as layer thickness reduction (LTR). Therefore, two design strategies were involved. • Strategy-1: If the designer decides<M ^extension in service life, the TBR value can be obtained with the use of Equation (1). NL TBR = N, (- \'B £V-R \EV-U ) where: N is the number of wheel load passes for a pavement surface deformation (rutting) up to the allowable rut depth in (mm). £v is the vertical compressive strain at the top of subgrade that can be obtained through the FE modeling Symbols R and U denote reinforced and unreinforced pavement sections. • Strategy-2:If the designer decides to keep the same service life of the reinforced section as it is for unreinforced section, the thickness of base/subbase may be reduced. The LTR can be calculated by Equation (2). LTR =\00T" ~T" (2) Hi IV (1) Abstract __ where: Tu and Tr are the layer (base or subbase) thickness of the unreinforced and reinforced sections, respectively. Design charts relating the vertical compressive strain at the top of subgrade with the layers thickness were developed. Since, the initial cost of construction of a pavement will increase due to additional cost of geogrids, the economical evaluation was taken as an essential part of the design of geogrid reinforced flexible pavements. The present worth method was employed taking into consideration the initial construction costs as well as both major and regular routine maintenance costs. The optimum section for geogridreinforced pavement was one having the highest values of TBR or LTR; along with lowest life cycle cost. The AASHTO procedure for design of traditional flexible pavements was also modified for its use in the design of geogrid-reinforced flexible pavements. The contribution of ^ geogrid is illustrated in terms of improvement in both the structural number and the layer coefficient. A computer program has been written in FORTRAN 77 to obtain the values of structural number for different soil support values (Mr -values). A simple chart has been obtained relating the values of structural numbers with the number of load repetitions and resilient modulus. The modified AASHTO procedure indicated that laying a layer of geogrid in subgrade, subbase or base course can reduce the thickness of these layers and provide the designer with many alternatives.