Modeling the Effect of Filler Materials on Performance of Hot Mix Asphalt Using Genetic Programming

This study aims to address the effect of filler materials (passing no. 200 sieve) on performance of hot mix asphalt mixtures. The Marshall mix design procedure was followed to prepare laboratory specimens using various types of filler materials to determine the optimum asphalt content. Furthermore, three different gradations with various amount of filler contents were selected to estimate the performance of asphalt mixture in laboratory conditions in terms of Marshall stability and flow. It was observed that more filler content, increases the plasticity of the mix and thus the permanent deformation. Moreover, adding the hydrated lime as filler improves the functional properties of the asphalt mixtures. A nonlinear genetic programming algorithm was developed to predict and formulize the Marshall stability results using material characteristics. The prediction power of the employed soft computing technique deemed satisfactory as compared to the experimental results.