Artificial intelligence approach for energy and entropy analyses of NiFe2O4/H2O nanofluid flow in a tube with vortex generator

Abstract

The base purpose of the study is to examine the influence of vortex generator geometry and nanofluid on thermo-hydraulic and irreversibility behavior by using a numerical and predictable approach under a laminar flow regime. The selection of an original magnetic nanofluid, the handling of forced convection in a tube including a vortex generator with the artificial neural network approach, and the 2nd law analysis of thermodynamics reflect the novelty of the study. In this context, it was explored numerically for heat transfer and flow profiles of NiFe2O4/H2O flowing with 1% volume fraction in a tube with different vortex generator geometries. This study has been carried out with the solutions under constant heat flux conditions of 2000 W/m2 along the tube. The analyzes have been applied in the range of 500≤Re≤2000. The Laminar model and single-phase approach in all analyses have been taken into account. Heat convection coefficient, pressure drop, and entropy generation were analyzed for the smooth tube and tube with wave ratio (WR=h/w) of 2, 3, and 4. Levenberg-Marquardt as train algorithm and Learngdm and Tansig as transfer function was used as the MLP network model in the present study. As a result, The highest heat transfer ratio, pressure drop, frictional entropy, and total entropy values are obtained for the tube with a wave ratio of 2. The highest deviation rates between the predicted and numerical results for WT1 and WT2 cases (the supreme among all cases) have been seen as 4.67% and 1.73% for the heat convection coefficient and pressure drop values, respectively.