Multivariate experimental investigation and RSM modeling of kinematic viscosity behavior in different classes of base oils

Kinematic viscosity is one of the essential parameters governing the performance and durability of lubricating oils in internal combustion engines. Unlike most previous studies that focus on a single base oil or isolated variables, this work provides a comparative multivariate analysis of viscosity variation across mineral, semi-synthetic, and synthetic base oils under simultaneous formulation and degradation effects. The influences of additive concentration, fuel dilution, oxidation level, shear rate, and base-oil type on kinematic viscosity were quantified using Response surface methodology (RSM). Kinematic viscosity measurements were conducted at 40 °C in accordance with ASTM D445. Oxidation severity was characterized by the total acid number (TAN), measured according to ASTM D664 and normalized between fresh and severely oxidized oils. ANOVA confirmed strong model performance (R 2 = 0.9804) and identified fuel dilution (F-value = 383.2), oxidation level (F-value = 287.1), and additive concentration (F-value = 256.6) as the dominant factors affecting viscosity, while shear rate had a negligible influence. Sensitivity analysis and regression coefficients explicitly demonstrated that mineral oils exhibit the highest viscosity sensitivity to degradation and formulation effects, whereas synthetic oils show measurably greater viscosity stability under identical conditions. The developed RSM-based model provides a practical framework for predicting the evolution of viscosity and supporting lubricant formulation and condition-monitoring strategies in engine lubrication systems.