The recent mechanistic-empirical approaches for predicting the remaining life of flexible pavements are mainly based upon the predicted strains at the interfaces of different layers. To determine these strains, nondestructive testing techniques are utilized. Unfortunately, uncertainties in determining the strains may result in significant errors in the predicted remaining life. A number of major factors that contribute to these inaccuracies include the imprecise knowledge of thickness and Poisson's ratio of each pavement layer and the errors in measuring the loads and the deflections using a NDT device.
A methodology which accounts for these uncertainties in the assumed pavement parameters and measured responses is suggested herein. With this methodology the influence of these parameters on the calculation of the remaining life of the pavement can be quantified. The methodology is based on Monte Carlo simulation techniques and has been used to analyze four pavement sections, representing a wide range of highways from secondary to interstate. The results of the probabilistic analysis show that the variability in pavement parameters increases the probability of failure of the pavement.