Abstract
This study discusses a methodology for developing a performance-based quality control system for flexible pavement works. The procedure relies on linking measures of quality at production and construction stages of the work with the in-service performance of flexible pavements. For this purpose, different elements of data from flexible pavement sections of 30 Wisconsin projects are compiled in a single relational database to demonstrate the potential of this methodology. From the compiled database, seven projects are selected for the in-depth analysis of the pavement. The effects of production and construction quality variation on performance are investigated in this study by using two performance prediction models that are based on the mechanistic-empirical method and statistical regression method. The prediction results of these models compared with the actual field performance measurements. The falling weight deflectometer (FWD) tests results improved the certainty of this comparison. The study shows the limitations associated with the use of current modeling programs in predicting long-term field performance, particularly in incorporating the quality variations inherent with flexible pavements. The analysis proposes a path to developing performance-based specifications for pavement materials based on understanding the relationship between quality measures and recorded performance of existing pavements. The findings suggest implementing a georeferenced relational database architecture for the Department of Transportations would magnify the impacts of current quality control programs.