International Journal of Thermofluids (Aug 2023)
AI predicts MHD double-diffusive mixed convection and entropy generation in hybrid-nanofluids for different magnetic field inclination angles by ANN
Abstract
The goal of the current numerical study is to determine how different magnetic field inclination angles along with Reynolds numbers affect mixed convective heat and mass transfer as well as entropy generation in a lid-driven trapezoidal enclosure with two rotating cylinders inside. As governing fluids, several water-based nanofluids and hybrid nanofluids with fixed solid volume fractions (5%) are employed. The impacts of three different types of SWCNT-Cu-Al2O3-water hybrid nanofluids, each made up of different ratios of SWCNT, Cu, and Al2O3 nanoparticles in water, are seen in this work, along with the effects of SWCNT-water, Cu-water, and Al2O3-water nanofluids separately. To obtain the results in the form of average Nusselt number, average Sherwood number, average temperature, and Bejan number as output parameters inside the enclosure for different magnetic field inclination angles (0° ≤ λ ≤ 90°) and other governing parameter values, the governing Navier-Stokes, thermal energy, and mass conservation equations are solved using the Galerkin weighted residual finite element method through numerical simulation. Then the simulation data is used to develop a novel artificial neural network model for efficient prediction. To achieve the best outcomes for the output parameters, the optimal values of each of these input parameters are determined by both FEM and ANN and a comparative study between these two is conducted. Cu-Al2O3-water hybrid nanofluid is used to test the performance of the developed ANN model for new cases. For each type of fluid in the present framework, maximum heat and mass transfer often take place when the magnetic field is applied at an inclined angle with respect to the horizontal axis. A satisfactory accuracy is obtained for the developed novel ANN model while predicting respective results. For training and validation data, it predicts convective heat and mass transfer with 96.81% accuracy and average dimensionless temperature and Bejan number with 98.74% accuracy. For test data, it predicts convective heat and mass transfer with 97.03% accuracy and average dimensionless temperature and Bejan number with 99.17% accuracy.
