Response Surface Methodology Investigation of the Viscosity of Propylene Glycol (100)/Graphene Nanofluid to Determine the Optimal Conditions

Abstract

This work adopted response surface methodology (RSM) to analyze the behavior of a nanofluid based on propylene glycol. The laboratory conditions in this investigation involve a temperature range of 40°C–120°C and a weight percentage that varies from 0% to 0.5%. Initially viscosity was predicted using Redwood viscometer using the nanofluid solutions. To find the most accurate predictive model and generate an ideal solution, RSM was used. The current study was inspired by the lack of consistency among laboratory behavior and real-world applications and the statistical-mathematical analysis of modelers' performance, contrast, and motivations. Two-factor interaction (2FI), quadratic, cubic, and quartic models are only a few tested. Investigating and evaluating the different statistical features of these modeling functions is a new contribution to the field. The quartic model represents the characteristics of nanofluids with double the accuracy of other models, as shown by statistical analysis. The R² coefficient, the coefficient of variation (CV%), and the p-value are compared as metrics for assessing the models. The indexes for the quartic model are 0.9940, 3.53%, and 0.0001, in that order. Nanofluids should have a viscosity of 0.335 m²/s at 120°C along with a weight percentage of 0.5%.