Permanent Deformation Predictive Equations Applicable to Mechanistic-Empirical Flexible Pavement Design

Abstract

Flexible pavement design is increasingly utilizing mechanistic-empirical (M-E) based design methods. To date, research has resulted in the development of two primary prediction models for permanent deformation on the basis of two different mechanistic-behavioral theories. An axially based model assumes that deformation is the result of axial compression, whereas a shear-based model assumes shear stresses within the asphalt layer are the primary cause of deformation. Permanent deformation is likely described by a model that incorporates both theories; however, little research exists in this area. The objective of this paper is to investigate numerous permanent deformation predictive models under the hypothesis that a model that combines both mechanistic theories reduces model form error. Improvement in the accuracy of performance predictions will allow designers to investigate novel mix designs, perform accurate reliability analyses, perform precise design optimization, and provide cost-effective designs for both initial construction and maintenance plans. Validation and comparison of numerous models through classical and Bayesian statistical methods will be achieved utilizing empirical data from the National Cooperative Highway Research Program’s (NCHRP) WesTrack project.