Development of a Framework for the Calibration of RPAS, a 3D Finite Element Analysis Tool for Rigid Pavements

The accurate analysis of rigid pavements requires a reliable modeling procedure based on integrating mechanistic analysis methods (i.e. closed-form solutions or numerical methods) and empirical observations (i.e. field measurements and laboratory test results). The use of the finite element method to model the response of rigid pavements has increased in recent decades due to its capability to incorporate the complexity of material behavior, traffic information, and environmental condition. Researchers from the University of Texas at El Paso developed the software Rigid Pavement Analysis System (RPAS) to comprehensively analyze the response of concrete pavements under different geometric configurations, foundation models, temperature gradient profiles and traffic loads by using the finite element method. Despite a few comparative studies that have been carried out during the development of this program, the implemented models and approaches may need improvement through a well-established calibration process. Therefore, this research aims to calibrate RPAS through a comparison with analytical solutions and field measurements. At the early stage of this study, a pre-validation was conducted in which field pavement critical responses and laboratory tests were compared with the responses predicted by RPAS. While a reasonable agreement between the responses was observed, it is the goal of the work presented here to develop and implement a calibration process that reduces the existing discrepancies. To this end, a series of studies that included verification and validation were conducted on a variety of pavement sections under different loading conditions. The calibration process of RPAS utilizes a multi-objective optimization algorithm that produces a list of calibration factors that are applied to the foundation moduli which was found, through a sensitivity study, to have a significant impact on pavement responses and also large variability. The calibration factors obtained ranged from 0.75 to 1.60 with most being close to 1.00 which, given the high variability in the foundation moduli, confirms the capability of RPAS in predicting the pavement responses with a good accuracy. The accuracy of RPAS after applying the calibration factors was assessed using a reliability metric that indicates a successful calibration process that brings the results produced by RPAS to within engineering expected thresholds. Ultimately, this research provides transportation agencies and pavement design engineers with a more reliable tool for the analysis of concrete pavements in comparison with the existing analysis tools