| The falling weight deflectometer (FWD) has become a popular tool worldwide for the evaluation of the structural capacity and integrity of existing pavements. Most of the deflection analysis programs used today to analyze the FWD deflections are based on static analysis, which often underestimates the subgrade strength. Unfortunately, dynamic analysis usually involves complex calculations and requires significant computation time thus making it impractical for routine applications.; This study presents a methodology based on deflection basin parameters and artificial neural networks (ANNs) for processing dynamic FWD measurements to estimate layer strengths. Two-dimensional, dynamic finite element analysis using the ABAQUS program was employed to develop the deflection information for this study. Unlike the majority of the existing backcalculation programs that iteratively adjust the layer moduli to match the measured deflections, the proposed method first determines the subgrade modulus by means of two deflection basin parameters, Base Damage Index and Shape Factor F2, and then applies the estimated subgrade modulus and other parameters as input variables to a trained ANN to estimate the upper layers' moduli. Procedures in predicting layer moduli for both two- and three-layer pavement systems are presented. In contrast to other programs that require the input of seed values for layer moduli, this method does not require initial estimates as input.; Field FWD measurements were analyzed both by this method and by the MODULUS program. Results reveal that the proposed method is better able to predict the asphalt concrete (AC) layer modulus while taking into account the dynamic effects of the FWD test. This method is also computationally efficient which makes it applicable for routine tasks and field use.; Effects of discontinuities in the AC layer of a pavement on the resulting deflections under an FWD test load were also studied using the finite element method. It was discovered that the condition of an AC layer, whether it is intact or damaged, may be detected using two deflection basin parameters, Shape Factor F2 and AREA. |
