RSM based modelling and optimization of Marshall properties of steel-slag and lime-modified asphalt mixtures

Abstract

Statistical and optimization techniques are useful tools in understanding the interactions and relationships between the asphalt mix variables and the contribution of each variable to the resulting asphalt mixture properties. This study explores the response surface methodology to study the influence of steel slag and lime on the Marshall properties of asphalt mixtures using Box Behnken Design tool. The independent variables considered are steel slag (0-100%), lime (0-4%) and bitumen (4-8%). The stability, flow, Marshall quotients, bulk density, Vb and VMA obtained were 1.98-6.35 kN, 3.27-4.53 mm, 0.53-1.60 kN/mm, 2.08-2.29 kg/m3, 7.67-14.7% and 14.02-24.63%, respectively; indicating that the steel-slag and lime-modified asphalt mixtures satisfy the specification limits recommended by the asphalt institute and Nigeria General Specification for Roads and Bridges. The models' analysis of variance revealed could well predict the Marshall properties of the mixtures, and the terms of steel-slag, lime and bitumen content are significant. Likewise, based on the optimization analysis, 24.93% of steel slag, 2.43% of lime and 5.51% of bitumen content were selected as the optimal values for the modified asphalt mixtures. Additionally, a mean error of less than 5% was attained for all the responses, demonstrating the effectiveness of RSM in designing asphalt mixtures.