Evaluation Of Pavement Me Locally Calibrated Distress Models For Jointed Plain Concrete Pavements In Pennsylvania

Abstract

The Pennsylvania Department of Transportation is currently in the process of implementing the Mechanistic Empirical Pavement Design Guide and its associated software AASHTOWare Pavement ME Design into its standard procedure for designing pavements. Among the key tasks of this implementation effort are the verification of the software's ``global'' pavement distress and smoothness prediction models and their local calibration to Pennsylvania conditions. These key tasks have been completed in a different study based on test sections located in Pennsylvania. This paper presents an evaluation of the locally calibrated pavement distress models for new Jointed Plain Concrete Pavements using three pavement sections outside the pool of test sections included in the verification and calibration tasks. Three approaches were used in the evaluation. The first two correspond to global and local models with field and laboratory measured input values, and the third one correspond to local coefficients with PennDOT default recommended input values. Additionally, 20-year designs were executed using Pavement ME and compared with designs completed using the 1993 AASHTO design guide. The results of this investigation indicated that for the sections evaluated the local calibration improved the prediction of joint faulting but had a negligible effect in the prediction of fatigue cracking. The global and local distress models under-predicted fatigue cracking and over-predicted faulting for the majority of cases evaluated. Additionally, it was found that the difference in slab thickness designs for Pavement ME and AASHTO93 tend to be larger when the input values deviate more from the local default recommended values.