Application of the logit model for the analysis of asphalt fatigue tests results

This paper explores the applicability of the logit function to reproduce the evolution of the stiffness of asphalt specimens during fatigue testing. Three logit-based models are formulated, and they are evaluated on the basis of a comprehensive database. Two of the models are proposed after such evaluation, so the most suitable one must be selected depending on the level of damage the specimen has undergone during testing (up to failure, beyond failure phase). One of these two models was also found to reproduce phase angle evolution. The general conclusion is that proposed models significantly improve performance of other models available so far (exponential, power, logarithmic, and Weibull), and provide an almost perfect fit to experimental data regardless of mixture type and testing procedure and conditions.