Estimation of density and moisture content in asphalt mixture based on dielectric property

•Propose a novel model termed ISO model for density prediction of asphalt mixture.•Clarify the effect of moisture on the dielectric property of asphalt mixture.•Quantitively estimate the moisture content in wet asphalt mixture. The density of asphalt mixture is a key quality property for in-situ pavement. And moisture within asphalt mixture easily results in various premature distresses on pavements. The basic goal of the present research is to investigate the estimation of both the density and moisture content within asphalt mixture using an electromagnetic (EM) technology. For this purpose, a generalized mixing formula was innovatively developed to predict the density of dry asphalt mixture regarding Influence coefficient v and Shape factor u based on EM mixing rules. Three types of asphalt mixture (AC-13, AC-20, and AC-25) specimens were fabricated in the laboratory with the same binder, asphalt content, and aggregate type, varying the density levels. Permittivity test of the specimens were conducted to obtain relative permittivity data. Using a nonlinear fitting method for the relationship between the bulk specific gravity and the relative permittivity of dry specimens, the optimal values of parameters v and u were determined to be 5.1 and −4.5 in the generalized mixing formula, defined as Influence and Shape Optimization (ISO) model. Moreover, considering the effect of water phase, a modified ISO (MISO) model was performed to estimate the moisture content of wet asphalt mixture. The main findings of the laboratory and field studies demonstrated that the ISO model was suitable for the density prediction of AC asphalt mixture within a reasonable error of 2%; and the MISO model predicted the moisture content with high R2 value not less than 0.74.