Evaluation of the performance of bio-oil modified asphalt mixtures through response surface methodology and optimization techniques

Abstract This study investigates the effect of bio-oil (BO), derived from sugarcane bagasse, and binder content (BC) on the volumetric and mechanical performance of bituminous concrete mixtures. BO was incorporated at four substitution levels (0, 3, 6, and 9%), while the BC ranged from 3.5 to 5.5%. The experimental design was developed using response surface methodology (RSM), and the data was analyzed with analysis of variance (ANOVA) to identify the most influential parameters. The results confirmed the suitability of a quadratic model, with a coefficient of determination R2 ranging from 0.70 to 0.99, indicating excellent agreement between the model and experimental data. Optimization, carried out using the desirability function based on the empirical model derived from RSM, allowed the determination of optimal conditions: a BC of 4.06% and a BO rate of 9%. Under these conditions, optimal values were obtained for several parameters: bulk density (BD = 2.35 g/cm³), theoretical density (TD = 2.46 g/cm³), air voids (AV = 3.54%), voids in mineral aggregate (VMA = 17.45%), voids filled with bitumen (VFB = 42.15%), unconditional Marshall stability (UCMS = 18.19 kN), conditioned Marshall stability (CMS = 17.86 kN), flow values (FV = 8.28 × 0.25 mm), Marshall quotient (MQ = 2.19 kN/mm), retained stability (RS = 97.56%), unconditioned indirect tensile strength (UCITS = 2244.74 kPa), conditioned indirect tensile strength (CITS = 2205.97 kPa), and tensile strength ratio (TSR = 97.66%). The optimization index achieved 65.5%, demonstrating that it is satisfactory despite the complexity of the criteria involved. In addition, analysis of the absolute relative error showed that it remained below 7%, confirming the reliability of the experimental measurements.