Modelling of 3D response pulse at the bottom of asphalt layer using a novel function and artificial neural network

Abstract

Fatigue life of asphalt mixes in laboratory tests is commonly determined by applying a haversine or sinusoidal load with a specific frequency. However, the frequency (time duration) and shape of horizontal tensile stress and strain pulses at the bottom the asphalt layer depend on pavement design (thickness and stiffness of layers) and loading specifications (speed and contact radius). The first objective of this paper is to introduce a novel function for a more realistic representation of 3D response (stress and strain pulses in longitudinal, transverse and vertical directions) at the bottom of the asphalt layer. The second objective is to establish a framework for determination of magnitude and shape of 3D response pulse at the bottom of the asphalt layer using artificial neural network. This framework enables designers to predict the shape and magnitude of stress and strain pulses in three directions based on some parameters related to pavement structure and loading specifications.