| The research developed procedures for assessing pavement layer condition using surface deflection data obtained from Falling Weight Deflectometer measurements. Both dynamic linear and nonlinear analyses were applied to pavement response modeling. The research seeks to improve the ability of assessing pavement layer conditions by identifying condition indicators, including deflection basin parameters, critical stresses and strains, and effective layer moduli, other than just using layer moduli. The procedures incorporate artificial neural network models as well as regression techniques for estimating condition indicators from surface deflections. Criteria for assessing layer conditions were developed by calibrating the procedure using field data with known layer material properties obtained from state DOTs and DataPave. The calibrated procedures were implemented into a Visual Basic computer program that has a user-friendly graphical interface.; The major strengths of the developed procedures are: (1) that the dynamic effect of FWD loading and nonlinear behavior of unbound materials were accounted for; (2) that the time to obtain results from the program is insignificant because these dynamic and nonlinear effects were included in the Artificial Neural Networks and regression equations; and (3) that some of the relationships constituting these procedures do not require all seven deflections, thus allowing the analysis of irregular deflection basins that occur frequently in distressed pavements for the layer condition assessment. |
