Integrated Taguchi-GRA-RSM optimization and ANN modelling of thermal performance of Zinc oxide nanofluids in an automobile Radiator

Abstract Impact of different input variables on the thermal performance features of an automobile radiator was investigated, statistically analyzed, and optimized using the powerful technique-Taguchi's grey relational analysis (GRA) and Response surface methodology (RSM). Polyethylene glycol (PEG) nanofluids containing ZnO nanoparticles of various volume concentrations (0.2%–0.6%) were used. 1-Butyl 3-methylimidazolium bromide [C4mim][Br] ionic liquid was added to reduce particle accumulation and increase nanofluid dispersion. The mini radiator used was an unmixed crossflow type. The analyses were carried out at various flow rates. Thermal performance parameters like Nusselt number (Nu), heat transfer coefficient (htc), pressure drop, and pumping power were optimized by using weighted grey relational grade, depending on the experiments designed using Taguchi's Experiment Design. ANN modelling was used to get a better prediction of the non-linear form of critical data. Optimized Nu, htc, pressure drop, and pumping power were obtained for different combination of Re, air velocity, and nanofluid concentration to maximize. The htc estimated by ANN is found to be reasonably consistent with the experimental findings. From the research findings, it is also inferred that heat transfer enhancement does occur in radiators employing nanofluids but at the expense of the pressure drop and pumping power.