FLEXIBLE PAVEMENT EVALUATION BY DEFLECTION CRITERIA

Abstract

Present oil crisis has resulted into enormous increase in the cost of bituminous road construction. 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 the existing ones. In India, like other developing countries, the problem of flexible pavement evalua tion is assuming greater significance as the existing pavements are subjected to continuous increase in traffic volume and axle loads. During the current sixth five year plan (1979-83),there is a massive rehabilitation programme in the road sector. Failure of flexible pavements is related to development of stresses and strains at critical points in a pavement under moving traffic. Measurement, of stresses/strains presents difficulty under field conditions. As far as evaluation of flexible pavement is considered deflection criteria are widely used because of the ease with which the deflections can be measured in the field with the conventional instruments. There is a need to establish adequacy of deflection as a criteria for pavement evaluation by studying the inter-relationships between the surface deflections and the stress/strain of the pavement system. Deflection parameters, essentially required for pavement evaluation are the maximum surface deflection and the deflection profile of the existing pavements. There is also a need to develop a. methodology to measure pavement response in terms of surface deflection under varying axle loads operative on the highway pavement system. To cater -11- for this need, a transfer function approach has been advanced in the present study, which utilizes the realistic in-situ input-output response of a pavement. If transfer function values are computed under standard axle load, they can be easily employed for the prediction of deflections under any spectrum of traffic loads. This would be helpful in the assess ment of the cumulative damage induced in a pavement due to fatigue. In order to establish the correlationship between two important parameters viz., the basic stress/strain and surface deflections of a pavement, a parametric study with a wide spectrum of variables is needed. For this purpose, a theoretical analysis of a layered pavement is imperative. For this part of the study, Finite Element Method has been used as theoreti cal tool for the pavement analysis. Finally, a methodology is suggested for the structural evaluation and overlay design of flexible pavements. The FIRST PART of this thesis covers basic concepts of a transfer function. The methods of determining transfer functions have been examined for first and second order systems using step and pulse load input. Transfer function is determined by using Laplace transformation, after assuming a suitable curve for the deflection output. For higher order system, it has been possible to develop an algorithm of z-transfer function which utilizes in-situ discretized input-output data. Transfer function in such a case involves solution of single input-output matrix, which can be computed without resorting -in to any complex curve fitting process. Prediction theory using discrete convolution is also presented. The SECOND PART of the thesis deals with experimental validation of transfer function on simplified versions of pavement systems, viz., beams of bituminous concrete on spring subgrade, a 2 layer pavement system and a 3-layer model scale pavement system.The study is also extended to a semi-full scale 3-layer pavement system. In these studies, time dependent load is taken as an input and time dependent deflections as an output. In the model study, the transfer function, is determined by using Laplace transformation with step load as input. For the semi-full scale pavement study, a random pulse load input is applied and both input and the corresponding output are simultaneously recorded and z-transfer function is determined directly by using the developed algorithm. Deflections predicted by transfer functions have been found to match fairly well with the experimental observations. In order to establish correlations between the basic stress/strain and the surface deflection of a pavement, a parametric study is essential. THIRD PART of the thesis is devoted to the analytical formulation of pavement structural response by finite element method, employing numerically integrated isoparametric quadrilateral elements. A computer programme 'MASTER ELAST' based on the above formulation is used. Finite element idealisation for a 3-layer pavement system, with 42 elements and I53 nodes, is arrived at after -ivcarrying out trials with different patterns and comparing the results with the known results through results of other investigations. Deflection profiles of semi-full scale pavement obtained from the plate load test and predicted by transfer function have been compared with theoretical predictions from FEM. The comparison revealed that surface deflection profile predicted by the use of transfer function is closer in shape and extent to the profile measured through the plate load test. In the FOURTH PART of the study, an attempt is made to examine the relationships between the basic stress/strain and surface deflection parameters of a pavement system. For this purpose, finite element model has been used. Inter relationships among maximum deflection, curvature, tensile strain and curvature-tensile strain ratio for the wearing surface and the compressive stress/strain on the top of subgrade have been examined. It is observed that maximum deflection and curvature of a deflected basin can be fruit fully employed for the evaluation of a pavement. For pave ments with thin wearing courses, maximum deflection as well as curvature have been found to be important, while for pavements with thick wearing courses only maximum deflection, criterion is adequate for the pavement evaluation. FIFTH PART of the thesis deals with the application of above studies to develop a methodology for the structural evaluation of existing pavements. It consists of measurement of deflection parameters for the existing pavements to give -vvalues of maximum deflection and curvature representative of the field condition. FEM computer programme is used with approximate values of layer moduli and other parameters like thickness of layers, load magnitude and contact area representative of the field conditions. Moduli of layers are adjusted by using iterative process through FEM computer programme till the computed deflection profile matches satisfactorily with the observed deflection basin. Thus knowing the layer moduli through the results of the para metric study, deflection data are transformed into the corresponding strain values. For structurally adequate pave ments, the computed strain values should not exceed the limiting strain after considering the effect of repetitions of traffic loads and other environmental factors. The proposed methodology for pavement evaluation overlay design has been explained with the help of a flow chart. Few reported case studies have been analysed to evaluate their structural adequacy and overlay thicknesses. This present study indicates that the transfer function technique along with the layered system analysis using finite element computer programme can be employed for the structural evaluation of pavement. Such a procedure will prove to be useful in economic and proper maintenance programmes of highway systems.