Development and calibration of permanent deformation models

Abstract

As part of the first accelerated pavement tests completed in Costa Rica, laboratory permanent deformation tests were performed on asphalt mixture, granular base and soil samples from the test sections. The objective of this study was to calibrate prediction models using measured rutting data collected from instrumented flexible pavements. For the asphalt concrete the laboratory test was performed using unconfined cyclic loading at three different temperatures. Permanent deformation was found to be a function of the resilient strain, temperature and number of loading cycles. For the granular base and soil samples, the test was performed under repeated axial cyclic stress at different magnitudes and different confining stresses. Permanent deformation was found to be a function of the confining stress, deviator stress, moisture content, and number of loading cycles. Four instrumented flexible pavements were subjected to heavy vehicle simulator testing. Backcalculated moduli from multi depth deflectometers were introduced into an elastic multilayer system to obtain pavement responses. The number of equivalent standard axle load repetitions, the effective asphalt concrete mean temperature, and the in-place moisture content along with computed responses were used to calculate permanent deformation for the different layers. By comparing the measured rut depths with the predicted values from laboratory based models, the optimum combination of field calibration factors was determined so that the coefficient of variation was minimal by means of ordinary least squares.