STRUCTURAL AND FUNCTIONAL PERFORMANCE MODELING OF LOW VOLUME ROADS

Abstract

Low volume rural roads form a major part of the Indian road network, comprising 80% of the total road length. These roads are usually constructed as granular pavements with or without thin bituminous surfacing layer and the design of these pavements is done using an empirical approach with subgrade California Bearing Ratio (CBR) as the main input. The empirical procedures are generally valid for the conditions addressed in the original study and have a tendency to introduce a level of uncertainty when extrapolated to other conditions. For the development of a rational pavement design criterion, development of performance models correlating structural and functional parameters of in-service pavements is essential. Low volume roads in India are currently designed based on analytical or theoretical solutions as per the guidelines given in Special Publication of Indian Roads Congress, IRC:SP:20 (2002) and IRC:SP:72 (2007). However, these documents do not refer to any mechanistic principles or to any documented performance data on which they are based. Therefore, a study needs to be done to evaluate the performance of different types of pavement under varying climatic and traffic conditions. It is necessary for reliable design of new roads and for selecting appropriate rehabilitation measures for in-service pavements. Rational design of low volume roads in India requires the development of performance evaluation criteria. For this purpose, this study was taken up with some selected road test sections in the northern part of India to gather experience on the performance of such roads under varying subgrade and traffic conditions and to develop a mechanistic-empirical performance criterion for low volume roads. The selection of the roads built under Prime Minister's rural connectivity programme was done based on various criteria, namely traffic conditions, surrounding environmental conditions, population base of the villages connected through the road under consideration, subgrade strengths and pavement thicknesses. On the whole, twenty road sections were identified from districts of Uttar Pradesh state and Uttarakhand state. DPRs (Detailed Project Reports) were examined to identify the secondary data on population size, rainfall intensity in the area, the iii section of the road, the CBR value of the subgrade soil, the depth of water table and types of aggregates used in the layers of a pavement in a road. Structural evaluation of pavement surface was carried out periodically with the help of Benkelman beam and Dynamic Cone Penetrometer (DCP). Average increase of 79% was observed for characteristic deflection and rut depth. Functional evaluation in terms- of roughness, longitudinal depression and other visual distresses was also evaluated. Average increase of 26% and 30% was observed for roughness and longitudinal roughness respectively. 2 sections each were observed "very good" and "poor" in rating of ride quality survey which was carried out on the test sections with the help of the experts. Remaining sections were found to be in "good" and "fair" conditions. Traffic (in terms of CVPD i.e. Commercial Vehicle Per Day) was found to have increased by 109% during the study period. Laboratory investigations were carried out for the material which was collected from the test pit evaluation on the test sections. Sieve analysis of granular material was performed to check the gradation of the base and sub-base materials. It was observed that except for some cases the gradation were matching the exact gradation as given in IRC:SP-20 (2002). Classification of the subgrade soil was done as per Indian Standards (IS) and American Association of State Highway and Transportation Officials (AAASHTO) classification. Atterberg limits, Optimum Moisture Content (OMC), Maximum Dry Density (MDD), Field Moisture Content (FMC), Field Dry Density (FDD) and degree of compaction were also calculated for subgrade soils. CBR, shear strength parameters (i.e. c and cp) and resilient modulus were calculated for the subgrade soils for OMC-FDD, FMC-FDD and OMC-MDD conditions. Material characterization was done with the help of the observed field and laboratory results. Relations were developed between CBR at field conditions, DCP index and resilient modulus at field conditions, and were compared with the results given by previous researchers. Power relationship was fitted well. It was found that CBR and MR decreased with increase in DCP index value. Relationships were also developed between CBR and MR values at field and laboratory conditions. For both iv the cases, power function was found to be the best fit and MR was found increasing with the increase in CBR value. Modified Structural Number (MSN) was also obtained with the help of UK DCP software. Pavement deterioration modeling was also done to relate the pavement distress parameters with its affecting factors like age, traffic, thickness of granular layer, subgrade properties, etc. Models were developed for deflection, roughness and rut depth. Linear and nonlinear models were developed with seventeen data sets and the final selected model was validated with the remaining 3 data sets. Nonlinear models were found to be best fit on the basis of logical relationship and statistical parameters. Relationship was also developed between the Present Serviceability Rating (PSR) and roughness values. Negative exponential relationship was found to be fitting well and as the value of roughness increased, PSR value decreased. Pavement responses (like vertical subgrade stress, vertical subgrade strain and surface deflections) were obtained from the Finite Element (FE) analysis of the 3-D model developed in the present study. Analysis was done considering the material to behave linearly and nonlinearly. Nonlinearity was modeled in different ways and applied in different layers. Significant difference was noted in the critical responses obtained from the linear and nonlinear (i.e. nonlinearity in granular layer) approach. The maximum increase in the case of vertical strain is 104% (54% in subgrade), while it is up to 45% in the case of surface deflection (at the axis of symmetry) as compared to linear analysis. When nonlinearity was modeled in subgrade, strain value became 25% higher on the interface of the granular and the subgrade layer. Surface deflections decreased as compared to linear analysis. The decrement was 8% to 19% as it moves radially outwards. When nonlinearity was modeled in both the layers i.e. granular and subgrade layer, the surface deflection and vertical subgrade strain increased by 37% to 59% respectively with respect to those obtained using linear analysis. Empirical relationship was also developed for estimating the equivalent linear elastic modulus of granular layer from modulus of the subgrade and the thickness of the granular layer. The equivalent linear elastic modulus estimated using this relationship increased with an increase in the subgrade strength and decreased with the increase in the thickness of the layer. u The pavements of the selected test sections were analyzed using three-dimensional finite element model that accounts for the nonlinear behavior of unbound pavement materials. Based on a riding quality survey, rut depth of 25 mm and International Roughness Index (IRI) of 9 m/km were identified as the terminal conditions for thin surfaced low traffic volume roads. Rutting and roughness trends obtained for the test sections were extrapolated to estimate the pavement lives in terms of cumulative standard axle load repetitions and number of years corresponding to each terminal condition. A mechanistic-empirical performance criterion was developed correlating the pavement life with vertical subgrade strain. Thickness design charts were prepared based on the subgrade strain criterion developed in the present study for the granular pavements with thin surfacing. The thickness obtained from this chart was compared with the chart given in IRC:SP:20 (2002) and IRC:SP:72 (2007). For IRC:SP:20 (2002) it was observed that for design traffic upto 0.3 million standard axle (msa) and for lower CBR (< 5%), lower thicknesses were obtained from the developed design chart as compared with the thickness provided in the IRC:SP:20 (2002). It was also observed that sections with high CBR value (mostly hilly sections) were provided with lower value of granular layer thickness, as compared to the granular thickness observed from the developed design chart. Modified thickness design chart was developed and compared with the thickness design chart as given in IRC:SP:20 (2002) and IRC:SP:72 (2007). The developed thickness design charts can be readily and easily used by the field engineers for estimation of granular layer thickness for a given strength of subgrade and traffic volume for roads catering traffic up to 1 msa. Similarly, relationships developed between material characteristics will be helpful in conversion of known material property to another. The developed deterioration models can be used to formulate maintenance management plans for low traffic volume roads. Keywords: Low traffic volume roads, Pavement performance, Deterioration models, Finite element analysis, Thickness design charts. vi