PERFORMANCE ANALYSIS OF OVERLAYS FOR FLEXIBLE PAVEMENTS

Abstract

In India, bituminous pavement surfacing are becoming increasingly common with the rapid growth of automobile traffic, and demand of the maintenance of existing road network. There is also a gradual tendency towards heavy pay-load vehicles and a general shift in emphasis from new construction of roads to the maintenance of existing ones. Very limited work has been done so far on the structural design efficacy of overlays With inherent deficiencies and various limitations, basic CBR method and its modifications are still used for the design of overlays on existing weak pavements in India. FIRST PART of the thesis forms the characterization tests in the laboratory on specimen and semi-full scale (SFS) testing on pavement sections. Characterization tests have been conducted in the laboratory for the identification of elastic constants and creep compliance of paving materials. Tests have been conducted on four types of the soil subgrade commonly found and representing the subgrade in pavements on Indian roads. These soils are sandy (SM), sandy clay (SO silty clay (SI.) and clay (CL). WBM mix employing crushed stone aggregate (CSA) representing base course materials and different types of bituminous specifications viz. bituminous concrete (BC), bituminous macadam (BM) and semi-dense carpet (SDC) representing pavement surface courses. Regression analysis of experimental data has resulted into development of empirical relationships indicating stress sensitivity of elastic modulii of constituent materials. To simulate horizontal tension in the bituminous layer, a test method has been employed using thick hollow cylinder triaxial specimens subjected to internal radial pressure and tested under axial compression. An empirical relationship has been developed to indicate the variation of stiffness modulus with internal radial and a: rial (!) pressures applied. The above tests have been conducted at a constant temperature of 30 ±1° Cwhich is arepresentative average temperature in India. Creep test have been conducted under uniaxial compression as well as tension, to obtain viscoelastic response of bituminous mixes under sustained loading conditions. Using experimental data, relationship have been developed to characterize the variation of creep compliance with sustained stress levels and duration of loading. To study the load deformation characteristics of various pavement systems employing different types of bituminous surface courses, semi-full scale pavement testing is done. The casting and testing of semi-full scale pavement sections includes the various thicknesses of WBM base courses and different types and thicknesses of bituminous surface courses (BC, BM, SDC) on compacted subgrade of four type of soils. In all, 16 simulated pavement systems have been tested. The pavement sections were adequately instrumented to measure the maximum surface deflection and surface deflection profile. SECOND PART (Chapter 3) of the thesis forms the development of models and relationships. An analytical model (FEP-PAO) has been developed for evaluation and design of overlays for flexible pavements. Finite element technique which is considered more powerful and reliable has been adopted for evaluating the performance of overlays for flexible pavements. Computer programmes have been developed and used for the pavement analysis for linearly, elastic, nonlinearly elastic and linear viscoelastic cases. These models are designated as FEPLE, FEP-PAO and FEPVIS respectively. The stiffness values of the pavement materials as determined from laboratory characterization tests (from the first part) are used as design input to the analytical model (FEP-PAO) to predict the load response characteristics of the semi-full scale pavement sections. The details of the development of relationships, overlay life prediction models and structural and roughness deterioration models have also been presented in this part. Details of the data collected from the overlaid flexible pavements and presentation of the data form the THIRD PART of the study (Chapter 4). The extensive field studies include the measurement of surface deflection, rut depth, roughness, cracks and cracking pattern and traffic volume. The data has been collected two times in year (i.e before the on-set monsoon and after the monsoon period is over) from 1988 to 1993. The field data also include the details of the temperature and moisture content measured during deflection studies, details of the soil subgrade, base courses and wearing courses. The measured data has been presented in Tables and Figures. The analysis and interpretation of results have been presented in Chapter 5 and form the FOURTH PART of the study. The analysis of the experimental data and the use of the developed model FEP- PAO Cmpredictions have been discussed in this part. FEP-PAO, a non-linearelastic model exhibits closer agreement with the experimental data than the conventional linear elastic model (FEPLE). FEP-PAO model has been used to develop equivalency values based on the criteria of equal surface deflection parameters (ESDP). There is a closer indication of better structural response of BC over BM + Six:. Details of the variation of influencing parameters (like characteristic deflection, Roughness Index (RI), Rut depth index (RDI), Crack index (CI) maintenance cost and traffic volume), with time is also presented in this part. These variations, form the basis for- taking maintenance decisions for further overlays. At any point of time, when any one of the influencing parameters reaches to its max'. mum (iii) acceptable limits, the need of overlay arises. Overlay life prediction models have been developed utilizing the data of existing pavement thickness, overlay thickness (ho), deflection (Dc), RDI, CI, RI and traffic volume (v). The models have been developed for each type of the soil subgrade. The structural and roughness deterioration models have been developed utilizing the data of the overlaid test sections. Calibration and validation of the models form the FIFTH PART of the study and are presented in Chapter Six. The developed model FEP-PAO has been calibrated using intensive field data collected from the Northern region of India. The model has been calibrated to yield maximum deflection equal to mean value (x) plus one times the standard deviation (<r) i.e. (x+tr). Trial run with the model indicates that the overlay thickness is adequate to keep the strain values within permissible limits. Validation of the model has been made by constructing the computed overlays of known thicknesses and remeasurement of rebound deflection and curvature in the field and checking them with the values obtained from the calibrated model. The measurement of rebound deflection and curvature have been made on overlaid pavements of known thicknesses. It has been observed that there is a very good agreement with the observed and the computed alues of deflection (upto 3 percent) and curvature (up to 10 percent). The overlay life prediction models have been calibrated and alldated through the intensive field data obtained through well designed extensive field programme. The calibrated model yields life of over-lay in months. Overlay thickness in terms of BM + SDC and PMC was constructed. The; continuous record of the various parameters (like Dc, h, ho, RDI, V, CI & RI) on the overlaid test sections were obtained. It was found satisfactory that the variation of up to 10 percent were recorded between the actual (observed) and computed life (iv) V V of the overlays. It gives ameasure of validation of the over-lay life prediction models developed for different conditions of terrain, subgrade soil, traffic and climatic conditions. The structural deterioration model has been calibrated to yield characteristic deflection (Dc) equal to mean deflection (x) plus one times the standard deviation (<r). The calibrated model yielded the deflection under the passes of comulative number of standard axles. The roughness deterioration model developed have been calibrated and validated through the intensive field data collected from extensive field studies on the overlaid flexible pavements located in the state of U.P. and H.P. The calibrated model yielded the roughness on the overlaid flexible pavements for a given number of cumulative number of standard axles. It has been observed that the selected existing overlaid flexible pavements require further overlays in terms of BM ♦ SDC and PMC to bring the Dc and roughness below the permissible limits. It was considered appropriate that the type and thicknesses of predetermined and designed overlays be constructed and a continuous record of commulative number of standard axles, surface deflection and roughness be made. It was found satisfactory that the variation of up to 3 percent and 10 percent between the actual (observed) and computed values of deflection and roughness respectively on the overlaid flexible pavements. It gives a measure of the validation of the structural and roughness deter-ioral Ion models developed for different conditions of soil subgrade. terrain and climatic conditions. The application of the developed models and relationships form the SIXTH PART of the study and are presented in Chapter Seven. The methodology as developed using FEP-PAO model is helpful in decision-making regarding the selection of the economical (v) pavement sections for overlays for maintenance of existing weak flexible pavements. The equivalency factors have been obtained using the FEP-PAO model by equalizing the surface deflection parameters (i.e. maximum surface deflection and curvature of deflection profile) for- BC and BM+SDC. Using FEP-PAO model, the overlay design charts have also been developed between surface deflection and different types and thicknesses of overlay in terms of BC and BM+SDC. The cost effectiveness study have been carried out through a case study by taking few existing pavements of commonly used thicknesses and soil subgrades in India. The relationships developed between deflection and RDI, CI and RI have been found to be useful for predicting the structural condition utilizing the data of functionl variables. The overlay life prediction models developed in this study found to be useful for predicting the life of overlays. Economic analysis has also been carried out for the selection of type and thicknesses of overlays °ut of tne overdesigned, standard designed and under designed thickness of overlays. Life of overlay per unit cost (months/Rs. ) has been determined through a case study by taking the data of few test secti ons It is hoped that the developed methodology in the form of analytical model (FEP-PAO), overlay life prediction models, relationship between deflection & rutdepth, cracking and roughness, relationship between deflection and rutdepth and ^ti'uctural and roughness deterioration models will be quite helpful to the pavement designers and field engineers, a like, for the selection of appropriate bituminous materials in view of the exhorbitant material cost, and the global need of developing appropriate rial Ion-based maintenance-management systems. (