Artificial neural network modeling for predicting surface texture and its attenuation of micro-surfacing containing steel slag aggregates

Abstract

• Artificial neural network models of micro-surfacing texture were established. • The effect of various factors on original texture and its attenuation were evaluated. • The traffic cycles and aggregate morphology has significant impact on surface texture. • Steel slag could improve the original texture and delay the attenuation of texture. For the research gap that the texture and its attenuation laws of micro-surfacing containing steel slag have not been considered, the objectives of this research were to establish the relationship between long-term texture features and engineering parameters using artificial neural network, and to determine the influence degree of each parameter. The results show that gradation and angularity are the most influential inputs for macro and micro-texture, respectively. The superior wear resistance of steel slag can delay the attenuation of micro-texture. The results of this study could provide alternatives to predict surface texture and offer sustainable way to recycle steel slag.