Statistical image analysis of uniformity of hybrid nanofluids and prediction models of thermophysical parameters based on artificial neural network (ANN)

Abstract

A new parameter (uniformity coefficient of nanoparticles distribution (UCND)) was introduced for quantitatively estimating the uniformity of nanoparticles distribution in nanofluids. Comparisons between UCND and measurements of the size distribution of nanoparticles show that the average diameters of nanoparticles in nanofluids (46:54, 50:50, 52:48, and 54:46) are 43, 54, 52, and 39 nm and the UCND are 0.9251, 0.9102, 0.9275, and 0.9513. They indicate that a mixture ratio of 54:46 is the most uniform with the lowest average diameter of nanoparticles and highest UCND. Moreover, nanofluids with the greatest uniformity exhibit the highest thermal conductivity with 0.426 W/(m.K) and relatively low viscosity with 3.761 mPa.s at 54:46, which is highly desirable for engineering applications. Finally, ANN models describing the thermal conductivity and viscosity were developed. The data predicted with the ANN models was agree with the experimental data, with (coefficient of determination) R2 = 0.9846 and R2 = 0.9755 for thermal conductivity and viscosity.