A meticulous intelligent approach to predict thermal conductivity ratio of hybrid nanofluids for heat transfer applications

Abstract

Accurate estimation of thermal conductivity of hybrid nanofluids is crucial for various engineering and industrial applications. The main objective of this research is to accurately predict the thermal conductivity ratio of hybrid nanofluids in water, ethylene glycol and various volume percentages of ethylene glycol/water as the base fluid. We adopted a novel soft computing technique, namely locally weighted linear regression (LWLR). Furthermore, for better validation, the linear genetic programming (LGP), gene expression programming (GEP) models and numerous empirical correlations were compared with LWLR model. The model outperformed LGP (RMSE = 0.0259, R = 0.964) and GEP (RMSE = 0.0474, R = 0.865) at RMSE = 0.014 and R = 0.988 for the prediction of thermal conductivity of hybrid nanofluids as revealed through performance criteria. Sensitivity analysis showed that the volume fraction, temperature and size of the nanoparticles were the most effective factors among all the inputs in the order of importance. Also, the thermal conductivity of nanoparticle and the mixing ratio of water and ethylene glycol were identified as the least effective factors on the thermal conductivity ratio of hybrid nanofluids assessment.