Thermophysical Properties Evaluation of Water Based Al2O3/TiO2 Hybrid Nanofluids: Experimental and ANN predictions

A new type of nanofluid, Al<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>O<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> and TiO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> nanoparticles, were dispersed and stabilised into the water base fluid. In the present investigation, we have synthesised 50 wt.% Al<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>O<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> and TiO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> nanoparticles. The prepared hybrid nanofluids were used in examining the thermophysical characteristics of the hybrid nanofluid. Experiment investigation of thermophysical characteristics was carried out at particle concentrations of 0.25–1.0 wt% and temperatures ranging from 20 to 60 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">o</sup>C. These were then estimated as three-dimensional thermophysical properties of the nanofluid using the Artificial Neural Network (ANN) Scaled Conjugate Gradient (SCG) approach. Hence, it is realised that the thermal conductivity of the hybrid nanofluids based on the present work is raised by about 42.73% at 60 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">o</sup>C onwards and the viscosity is increased by about 68.73% at 20 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">o</sup>C for the base fluid. In fact, the density of hybrid nanofluid is increases with the addition of nanoparticles; in the same way, the specific heat capacity of hybrid nanofluids decreases with the addition of nanoparticles. The experimentally evaluated property data is compared with the literature data. High-accurate predicted experimental data is obtained by using the ANN-SCG algorithm. Thus, using the ANN-SCG, the trained data has the following: coefficient of correction of thermal conductivity = 0.9777, the coefficient of correction of viscosity = 0.9983, that of the specific heat = 0.9891, and that of density = the mean squares error = (0.00001,1.3346), respectively. 0.0003, 3.4668, and 1.3346), respectively. Based on the experimental results, an equation of the general regression form was obtained.