This study is concerned with enhancing the reliability analysis currently used in the Mechanistic-Empirical Pavement Design Guide. The Mechanistic-Empirical Pavement Design Guide uses a set of mechanistic-empirical models to analyze distresses of pavement structures in response to traffic, climate, and materials. These models include transverse joint faulting and transverse cracking for rigid pavements, and permanent deformation, fatigue cracking, and thermal cracking for flexible pavements. In order to evaluate the reliability of a selected pavement structure, the current procedure utilizes the overall standard deviation of the measured distresses obtained from calibration against distressed pavements in comparison with predicted values. This technique is fairly simple; however, it is far from accurate. Some limitations of the method may be attributed to the standard deviation being calibration-site specific, as well as the assumption of normality of the distribution of the measured distress levels. Additionally, this method relies on the variability of the measured output rather than on the variability/uncertainty of the input parameters that induce such variability in distress. The identification of an improved methodology for assessing design reliability is a top priority. In this study, the uncertainties in model input parameters (materials, climate, traffic) due to inherent spatial variability of materials and testing nonuniformity will be incorporated by using a number of advanced reliability techniques to assess the reliability of flexible and rigid pavements. Both analytical and simulation will be investigated. Materials variability and uncertainty associated with estimating many other input parameters will be determined from historical records. The reliability associated with pavement smoothness, as a primary measure of ride quality, in response to the various distresses encountered will be analyzed. Design reliabilities obtained from the various methods used will be compared to the current method. Advantages and disadvantages of the new methodologies in comparison with the current method will be discussed in light of both the accuracy and computational feasibility or practicality.
The key project objectives are:
- Reliable and effective pavement design.
- Efficient reliability analysis for pavement design.
The research findings have been presented in three peer-reviewed journal papers. The research developed an efficient reliability procedure applicable to the National Cooperative Highway Research Program 1-37A Mechanistic-Empirical Pavement Design Guide and statistical analysis procedures for the identification of critical project-specific input parameters.