DESIGN ANALYSIS AND CHARACTERISATION OF FLEXIBLE PAVEMENTS

Abstract

Several kinds of bituminous pavement specifications are in vogue for the trunk routes and urban arterial systems in Indie. These include bituminous concrete (BC), bitumen bound macadam (BBM), sheet asphalt (SA) etc. and are based on the recommended practices of the Indian Roads Congress (IRC). However, very limited work has been done so far on structural characterisation of bituminous pavement layers and as such IRC specifications are absolutely quiet about it. In fact, the design of flexible pavement system in India is based on adhoc experience and empirical strength tests such as the California Bearing Ratio (CBR) test. This study is planned to develop suitable analytical model to analyse the load responses of bituminous pavements based on more realistic strength characterisation tests simulating the stress environment in pavement components under wheel loads. The study can be divided into four parts. The FIRST PART aims at developing a mechanistic approach to analyse bituminous pavements. Finite element technique has been adopted for the analysis to account for the linearly elastic, nonlinearly elastic and viscoelastic behaviour of pavement components under traffic loads. Computer programmes have been developed for use on DEC 2050 Computer of the Regional Computer Centre of the University of Roorkee (UOR). The above programmes are designated as UORFEP (IE), UORFEP (NLE) and UORFEP (VIS), respectively. ii In each case the finite element model of width and depth 12 times the radius of loading area, has been chosen with axisymmetric loading. Isoparametric quadrilateralelements have been adopted for the analysis of multi-layer pavement system in which the interfaces are assumed to be perfectly rough. The selection of number of elements and nodal points has been done taking into consideration the compatibility requirements without sacrificing accuracy. The analysis is started with pseudo elastic constants for elements in each layer of the continuum. The stress compo nents and principal stresses are obtained for the given pavement system and loading configuration. With principal stresses evolved for each element and with the known stress oduli constitutive relationships developed from laboratory characterisation tests, a new set of strength moduli are calculated for each element for use in the subsequent cycle of solution. The process is done automatically and sequentially on the digital computer. The interactive process is continued until difference between the previous and new values of moduli for each element is less than 0.5 percent. The SECOND PART of the study deals with material characterisation tests which would form the necessary inputs into the analytical models developed as mentioned above. Tests have been conducted on local soil (IS Classification - SM) which represents the subgrade, WBM mix comprising crushed m Ill stone aggregate (CSA)/overburnt brick aggregate (OBA) which form the base courses and the bituminous specifications, viz. BC, BBM and SA forming the surface courses of the pavement system. Elastic characterisation has been modified to include stress sensitivity of paving materials by conducting the standard triaxial test at various levels of confining pressure. Tc simulate the horizontal tension developed at the bottom of the bituminous surface course, hollow cylinder triaxial test technique has been used in which the hollow cylindrical specimens are subjected to internal radial pressures and tested under axial compression. It has been possible to develop an empirical relationship between stiff ness modulus, internal radial pressure and axial stress. Creep tests in compression as well as in tension have been conducted on bituminous specifications to yield their viscoelastic responses under sustained loading conditions. Constitutive equations have been developed from experimentalresults to express the influence of time on creep compliance. Load response studies of test tracks and semi-full scale pavement sections have always drawn considerable interest of highway engineers and researchers to evaluate the pavement mechanistic models and distress models. The casting and testing of semi-full pavement sections employing the various combinations of WBM base courses and bituminous iv surface courses form the THIRD PART of the study. Pavement sections of horizontal dimensions 2.8m x 2.57m have been constructed in the pavement testing laboratory of the University. After subgrade has been compacted with local soil, 3 types of base courses, two of CSA having thicknesses 150 mm and 225 mm and the third of OBA with I50 mm thickness, were installed following the IRC recommended practice. BC of different thicknesses, viz. 25 mm, 50 mm and 75 mm and BBM and SA each of 50 mm and 75 mm were installed on each type of base course. In all 21 simulated pavement systems have been tested under static plate load of 200 mm diameter. Each pavement section was adequately instrumented to measure the surface deflection profile, reactive pressure at interfaces and moisture and temperature variations during testing. Load response test data in terms of maximum deflection, deflection profiles and curvatures and reactive pressures at interfaces under the load axis have been analysed to quantify the structural efficacy of various types of bituminous pavement components and the CSA and OBA base course. Analy tical predictions from UORFEP (IE) and UORFEP (NLE) models have also been made using the constitutive relationships of elastic constants developed in the second part of the study. The latter model is found to show closer agreement with the experimental observations. In order to study the relative efficacy of various types of bituminous specifications and layer thicknesses, an attempt has been made to develop equivalency values using UORFEP (HIE) model based on equal reactive pressure on the subgrade. Although great signifi cance cannot be attached to the equivalency value due to its inherent limitations, nevertheless there is a clear indication of better structural response of BC and BBM over SA in the order mentioned above. The FOURTH PART of the study summarises the application of the analytical model and material characterisation tests for design of new pavements and evaluation and strengthening of existing flexible pavements. Principal criteria for design in terms of limiting compressive strain in subgrade and tensile strain in bituminous surface course for expected applications of equivalent axle loads, have been summarised. The design procedures employing the above criteria have been explained by flow charts. A case study explaining the use of UORFEP (NLE) model for determination of overlay thick ness has also been included in this work. The above model has considerable potential in designing and strengthening of flexible pavements.