Performance evaluation of Bailey method used in asphalt mixtures containing natural river sands

Abstract

This study aims to investigate the efficiency of the Bailey method when natural sand is included in the mix at two coarseness levels of aggregate gradations: fine-graded (FG) and coarse-graided (CG), and three mixes with varing percentages of the natural river sand were prepared at each coarseness level, namely: CG with quarry sand only (CG-QS), CG with natural sands only (CG-NS), CG with quarry and natural sands (CG-QNS), FG with quarry sand only (FG-QS), FG with natural sands only (FG-NS), and FG with quarry and natural sands (CG-QNS). The portions of the natural sand either in CG-QNS and FG-QNS mixes were minimized as possible without violating the Bailey ratios. Four performance tests were utilized to evaluate the mixes including indirect tensile tension test (IDEAL-CT), disk‐shaped compact tension test (DCT), flow number (FN) test, and tensile strength ratio (TSR) test. The sensitivity and the significance of the volumetric measures were investigated including voids of total mix (VTM), voids of mineral aggregate (VMA), dust proportion (DP), voids filled with asphalt (VFA), air voids volume (V_a), and effective binder volume (V_be), and a number of non-linear statistical models were developed. The results of IDEAL-CT, DCT, and FN test were consistent where CG-QNS had the best performance, followed by CG-QS, CG-NS, FG-QS, FG-QNS, and lastly FG-NS. TSR value was always decreasing with the increase in the percentage of the natural river. V_be was the most significant volumetric measure followed and was used to predict the performance test indices. FN test had the highest coefficient of determination (R²) when modeled with any of the other performance tests, while TSR test had the lowest. The Bailey gradation method was successfully proficient to provide a similar gradation coarseness for CG-QNS when compared to CG-QS which indicates a similar aggregate interlock.