Investigation into stability and thermal-fluid behaviour of hybrid nanofluids as heat transfer fluids

Abstract

The need to improve the poor thermal conductivity of conventional fluids to produce adequate heat transfer fluid cannot be over-emphasized, knowing fully well that heat transfer is key in any engineering process line. Hence, the birth of nanofluids, which is the formulation of a composite of suspended nanoparticles in a basefluid. Nanofluids have found wide applications ranging from heat exchangers, electronic cooling, automotive industry, medical, military, solar energy, manufacturing industry, to mention but a few. But the limitations of nanofluids led to the entrance of a new working fluid named binary nanofluid and ternary nanofluid. This study experimented with the trio influence of temperature (T), percent weight ratios (PWRs), nanoparticles size (NS) on the thermophysical behaviour of MgO–ZnO/Deionised water binary nanofluids (BNFs). 20 nm nano-size of ZnO nanoparticles were hybridised with MgO nanoparticles of nano-sizes 20 nm and 100 nm, and dispersed in deionised water to prepare 0.1 vol% binary nanofluids for percent weight ratios of MgO:ZnO (20:80, 40:60, 60:40 and 80:20). The viscosity (μ), electrical conductivity (σ), pH, and thermal conductivity (κ) of the binary nanofluids were experimentally evaluated for temperature 20 to 50 °C. Morphology was checked, and stability was monitored. The impact of temperature, PWRs, and nano-size on the pH, μ, σ, and κ of the binary nanofluid were ordered as PWR >NS >T, NS> PWR>T, T>NS >PWR, and T >NS >PWR, respectively. Using the obtained experimental dataset, correlations were proposed for the thermal property of each binary nanofluid as a function of temperature. Also, investigating the trio impact of PWR, temperature and 𝜑� on the thermophysical characteristics of MgO-ZnO/DIW BNFs, to help close up the scarce literature gap. 20 nm nanoparticle sizes of MgO and ZnO were hybridized together and dissolved in deionized water to formulate 0.1 vol% and 0.05 vol.% binary nanofluids (NFs) for PWR of 20:80, 40:60, 60:40, 80:20 (MgO:ZnO). The κ for all BNFs was enhanced under the impact of rising temperature, with maximum κ enhancement of 5.60% and 22.07% relative to the deionised water (DIW) achieved for 0.05 vol% and 0.10 vol%, separately. The σ was enhanced slightly under the influence of increasing temperature, with maximum enhancement of 21.82% and 30.91% achieved for 0.050 vol% and 0.10 vol%, respectively. In addition, viscosity under temperature increase exhibited a decreasing pattern for all nanohybrids and basefluid. Furthermore, to better harness the benefit of the BNFs for thermal application, thermoelectrical conductivity (TEC) was evaluated with BNFs of 0.05 vol% observed to have higher TEC values than 0.10 vol% BNFs. The BNFs were found suitable as thermal fluids. A novel manner of furthering thermo-convection behaviour of thermal applications is the use of BNFs as heat transfer fluids. This study experimented the natural convection behaviour of MgO-ZnO NPs suspended in basefluid for 𝜑� = 0.050 vol.% and 0.10 vol% at percent weight ratios of 20:80, 40:60, 60:40, 80:20 (MgO:ZnO) inside a square enclosure. Factors like Rayleigh number, Nusselt number (Nuav), coefficient of convective heat transfer (hav), and heat transfer rate (Qav) for various temperatures (20°C to 50°C) were examined. PWRs and temperature gradient of BNPs inside the binary nanofluids was observed to augment Nuav, hav, and Qav. Also, highest improvement of 72.60% (Nuav), 76.01% (hav), and 72.20% (Qav) was achieved. Employing BNFs in square enclosure yielded fine improvement for natural convection behaviour. Artificial intelligence (AI) methods, like artificial neural network (ANN) and surface fitting method were deployed to model the thermal conductivity of BNFs. For the ANN model, a learning algorithm was developed to determine the optimum neuron number. The ANN having 19 neurons in the inner layer got the optimized performance. A surface fitting method was also used on the experimental data, and the generated surface shows the behaviour of the BNFs. The outcome affirmed that the designed ANN model is best for predicting the thermal conductivity of MgO-ZnO/DIW binary nanofluids for different temperatures, nanoparticle sizes, PWRs and volume concentration over the surface fitting method.