The Mechanistic-Empirical Pavement Design Guide (MEPDG) accounts for a large range of factors affecting the pavement performance such as foundation/subgrade, paving materials, environmental factors (temperature and moisture), traffic loadings, sub-drainage, shoulder design, pavement performance(distresses and smoothness), and design reliability. It should be noted, however, that the MEDPG software is a pavement analysis tool which can be used to predict pavement performance for selected design parameters but does not determine the design features, e.g. layer thicknesses, resulting inacceptable and economical design. A framework of DAKOTA and MEPDG has been developed to bridge this gap in facilitating the optimal design of pavements and is discussed herein. Two design examples of flexible pavements are considered. The cost of construction of a unit length and width segment of the pavement serves as the design goal, i.e. the objective function of the related optimization problem. Six constraints involving various designreliabilities are imposed during the optimization process to ensure that the pavement design satisfies functional requirements during its service life. Two types of optimization procedures are used: direct optimization and surrogate- based optimization. During the course of the study, a slight anomaly in the designs generated by the MEPDG was observed and is reported.
Samenvatting